Re: attempted install of buster arm64 net-install on rp4 fails instantly

[David Wright]

On Sun 08 Sep 2019 at 13:26:16 (-0400), Gene Heskett wrote:
> On Sunday 08 September 2019 12:49:14 Curt wrote:
> > On 2019-09-08, Gene Heskett  wrote:
> > >> > What does this do that the iso doesn't, and note it takes a
> > >> > windows machine to follow those instructions.
> > >>
> > >> Eh? How do you work that out?
> > >
> > > I won't, theres always another way to skin that cat. ;-) Time, if I
> > > have it, will sort that out.
> >
> > He means how did you derive that erroneous notion from the
> > 'README.concatenateable_images,' which says:
> >
> >  To create a complete image from the two parts on Linux systems, you
> > can use zcat as follows:
> >
> >zcat firmware..img.gz partition.img.gz >
> > complete_image.img
> 
> zcat is not a cmnd I use, so I didn't recognize it.

Perhaps you've been brainwashed by the "pointless use of cat" crowd :)

> Besides, it more of 
> the zero info names that one isn't going to remember a week down the log 
> when one trips over partition.img.gz and wonders why it was that I 
> downloaded it.

Easily fixed by placing them in directories with meaningful names,
as in:

debian/iso/buster/10.0.0/netboot/d-i/amd64/initrd.gz (2019-07-01 30581523)
debian/iso/buster/10.0.0/netboot/d-i/amd64/linux (2019-07-01 5221616)
debian/iso/buster/10.0.0/netboot/d-i/amd64/SHA256SUMS (2019-07-01 75295)
debian/iso/buster/10.0.0/netboot/d-i/i386/initrd.gz (2019-07-01 26665346)
debian/iso/buster/10.0.0/netboot/d-i/i386/linux (2019-07-01 4140688)
debian/iso/buster/10.0.0/netboot/d-i/i386/SHA256SUMS (2019-07-01 73988)
debian/iso/buster/10.0.0/netinst/amd64/firmware-10.0.0-amd64-netinst.iso 
(2019-07-06 393216000)
debian/iso/buster/10.0.0/netinst/amd64/SHA512SUMS (2019-07-06 332)
debian/iso/buster/10.0.0/netinst/amd64/SHA512SUMS.sign (2019-07-06 833)
debian/iso/buster/10.0.0/netinst/i386/firmware-10.0.0-i386-netinst.iso 
(2019-07-06 488636416)
debian/iso/buster/10.0.0/netinst/i386/SHA512SUMS (2019-07-06 330)
debian/iso/buster/10.0.0/netinst/i386/SHA512SUMS.sign (2019-07-06 833)
debian/iso/jessie/8.7.1/netinst/i386/firmware-8.7.1-i386-netinst.iso 
(2017-01-16 351272960)
debian/iso/jessie/8.7.1/netinst/i386/SHA512SUMS (2017-01-17 162)
debian/iso/jessie/8.7.1/netinst/i386/SHA512SUMS.sign (2017-01-17 819)
debian/iso/stretch/9.1.0/live/amd64/debian-live-9.1.0-amd64-lxde+nonfree.iso 
(2017-07-22 2346844160)
debian/iso/stretch/9.1.0/live/amd64/SHA512SUMS (2017-07-23 4180)
debian/iso/stretch/9.1.0/live/amd64/SHA512SUMS.sign (2017-07-23 833)
debian/iso/stretch/9.1.0/live/i386/debian-live-9.1.0-i386-lxde+nonfree.iso 
(2017-07-23 2394816512)
debian/iso/stretch/9.1.0/live/i386/SHA512SUMS (2017-07-23 4156)
debian/iso/stretch/9.1.0/live/i386/SHA512SUMS.sign (2017-07-23 833)
debian/iso/stretch/9.5.0/netinst/amd64/firmware-9.5.0-amd64-netinst.iso 
(2018-07-14 341835776)
debian/iso/stretch/9.5.0/netinst/amd64/SHA512SUMS (2018-07-14 163)
debian/iso/stretch/9.5.0/netinst/amd64/SHA512SUMS.sign (2018-07-14 833)
debian/iso/stretch/9.5.0/netinst/i386/firmware-9.5.0-i386-netinst.iso 
(2018-07-14 432013312)
debian/iso/stretch/9.5.0/netinst/i386/SHA512SUMS (2018-07-14 162)
debian/iso/stretch/9.5.0/netinst/i386/SHA512SUMS.sign (2018-07-14 833)

Cheers,
David.

Re: Cannot connect to HP printer

[Paul Duncan]

Have you tried using HPLIP? The only thing that I have not managed to get
working using that is a DesignJet T120.

On Sun, 8 Sep 2019, 10:14 Brian,  wrote:

> On Sat 07 Sep 2019 at 23:20:03 +0200, deloptes wrote:
>
> > Brian wrote:
> >
> > > Every printer that exists employs lpadmin with whatever utility is
> > > used to set it up. It is basic, whether or not the device is just a
> > > printer or an MFP. The hplip package is a convenient software but
> > > completely unneeded for printing.
> > >
> > > Don't confuse printing and scanning. They are separate processes.
> > > With an HP MFD I'd give you that the libsane-hpaio package is needed,
> > > but the hplip package is not required.
> > >
> > > (Actually, with a modern MFD, libsane-hpaio might not be required).
> >
> > Thanks, I have no visibility. I use a HP printer with cups and separate
> > Epson scanner with iscan.
> >
> > I know print and scan are separate things - well, it is obvious isn't it
> -
>
> Not that obvious, apparently. Many users start off by describing an issue
> with "My printer is working but by scanner is not". How does "My printer
> is working but by mouse is not" sound?
>
> It's just one of those things. Like thinking that hplip is required to
> get an HP printer operative on Debian.
>
> --
> Brian.
>
>

Re: Lost sound after update to 9:9.10

[Juan R. de Silva]

On Sun, 08 Sep 2019 20:42:46 +0200, Sven Joachim wrote:

> On 2019-09-08 18:13 +, Juan R. de Silva wrote:
> 
>> One of my computers lost sound (both speakers and headphones) after
>> update to 9:9.10. User is in "audio" group. No sound neither for the
>> user nor for root. Hardware tested with Live Ubuntu 16.04 is alive and
>> functioning.
>>
>> Any suggestions, please?
> 
> Try booting with the previous kernel, linux-image-4.9.0-9-amd64.  If
> that brings the sound back, file a bug against src:linux (preferably
> with reportbug).
> 
> Cheers,
>Sven

Hi Sven,

I've just get my sound back by installing Pavucontrol. It seems the 
problem was in GNOME All Settings Sound applet. For some reason in 
"Choose a device for sound output" it is listing only Digital Output (S/
PDIF) Built-in Audio device and showes only one  profile "Digital Stereo 
(IEC958) Output" available, which left me without any sound.

Pavucontrol showed all aptions available. I was able to select Show 
Hardware Output Devices and to select Analog Stereo Output (which BTW 
then appeared in the GNOME Sound applet as well). This get my sound back. 
However in the GNOME Sound applet I still have only Digital Output (S/
PDIF) Built-in Audio device listed. Thus if I by mistake click on that I 
loose Analog Stereo Output option again and have to launch Pavucontrol 
again to correct the problem.

I would report the problem as a bug. The problem is that I not yet fully 
understand what actually happens. Should I report a bug against GNOME 
Sound applet or the problem is laying anywhere else? Seems that otherwise 
my card is detected and configured correctly.

Thanks.

Re: Cannot connect to HP printer

[Brian]

On Sat 07 Sep 2019 at 23:20:03 +0200, deloptes wrote:

> Brian wrote:
> 
> > Every printer that exists employs lpadmin with whatever utility is
> > used to set it up. It is basic, whether or not the device is just a
> > printer or an MFP. The hplip package is a convenient software but
> > completely unneeded for printing.
> > 
> > Don't confuse printing and scanning. They are separate processes.
> > With an HP MFD I'd give you that the libsane-hpaio package is needed,
> > but the hplip package is not required.
> > 
> > (Actually, with a modern MFD, libsane-hpaio might not be required).
> 
> Thanks, I have no visibility. I use a HP printer with cups and separate
> Epson scanner with iscan.
> 
> I know print and scan are separate things - well, it is obvious isn't it -

Not that obvious, apparently. Many users start off by describing an issue
with "My printer is working but by scanner is not". How does "My printer
is working but by mouse is not" sound?

It's just one of those things. Like thinking that hplip is required to
get an HP printer operative on Debian.

-- 
Brian.

Linux Journal epub files

[Marc Shapiro]

I have Linux Journal in epub format up through June of 2018.  I didn't 
grab the others as they became available and by the time I realized that 
LJ was closing permanently, the archives were no longer working.  Does 
anyone know where I can find the newer issues (7/18 through end of 
publication) in epub format?



Marc

Re: attempted install of buster arm64 net-install on rp4 fails instantly

[Gene Heskett]

On Sunday 08 September 2019 13:46:11 Gene Heskett wrote:

> On Sunday 08 September 2019 11:34:20 Gene Heskett wrote:
> > On Sunday 08 September 2019 08:35:51 Thomas Schmitt wrote:
> > > Hi,
> > >
> > > i now realize that Gene by "rp4" indicated some known Raspberry
> > > system. So my proposal about netboot might be hopeless according
> > > to the answers of David and didier.gaumet.
> > >
> > > --
> > >-- -- But there is technical stuff left to discuss:
> > >
> > > I wrote:
> > > > > SD card seems to be the intended target for netboot images.
> > > > > http://ftp.debian.org/debian/dists/stable/main/installer-arm64
> > > > >/c ur rent/images/netboot/SD-card-images/ [...]
> > > > > Those are not ISO9660 but rather partitioned images with a
> > > > > FAT32 filesystem:
> > > > > [...]
> > > > >   Device  Boot StartEnd Sectors  Size
> > > > > Id Type firmware.a64-olinuxino.img1 * 2048 19  197952
> > > > > 96.7M c W95
> > >
> > > Gene Heskett wrote:
> > > > What does this do that the iso doesn't,
> > >
> > > Those are images for disk-like devices only and they don't look
> > > like they are supposed to boot directly via EFI. The Debian arm64
> > > ISOs on the other hand offer typical EFI boot equipment.
> > >
> > > > and note it takes a windows  machine to follow those
> > > > instructions.
> > >
> > > If you mean
> > >
> > > http://ftp.debian.org/debian/dists/stable/main/installer-arm64/cur
> > >re nt /images/netboot/SD-card-images/README.concatenateable_images
> > > then immediately before the MS-Windows instructions, i read:
> > >
> > >   "To create a complete image from the two parts on Linux systems,
> > > you can use zcat as follows:
> > >  zcat firmware..img.gz partition.img.gz >
> > > complete_image.img "
> > > I inspected both image parts by gunzip-ing them:
> > >
> > >   $ gunzip firmware.a64-olinuxino.img.gz
> > >   $ /sbin/fdisk -lu firmware.a64-olinuxino.img
> > >   ... lots of message lines ...
> > >   $ gunzip partition.img.gz
> > >   $ sudo mount partition.img /mnt/fat
> > >   $ find /mnt/fat | less
> > >
> > > So i would simply provide another useless use of cat with my
> > > superuser hat on:
> > >
> > >   # cat firmware.a64-olinuxino.img partition.img | dd bs=4096
> > > of=/dev/sdf1
> > >
> > > (Insert "sudo" where needed, if your system has no superuser.)
> > >
> > > > I think what I will do next is send gparted to create a gpt
> > > > table with a fat32 first partition of a gig or so, and do an
> > > > ext4 on the rest of it. Then put the iso on sdf.
> > >
> > > The ISO brings its own partition table. Your gparted work will be
> > > wasted. See the second grey box at
> > >   https://wiki.debian.org/RepackBootableISO#arm64_release_9.4.0
> > > which shows fdisk output for arm64 ISOs.
> > > (Nothing did change in the ISO partitioning for 10.0 since 9.4.)
> > >
> > > > Since these come as NTFS formatted cards these days,
> > >
> > > Filesystems on the card get overwritten by the ISO or at least
> > > lose their entry in the partition table, if they are not reached
> > > by dd's work.
> > >
> > > > I would think that
> > > > writing the iso to /dev/sdf would at least start the install,
> > > > and the installers disk utils could take care of the rest.  What
> > > > the iso wrote should be all that counts.
> > >
> > > As said, Debian arm64 netinst ISOs offer boot entries for EFI.
> > > Googling "raspberry rp4 efi" shows that the combination of
> > > Raspberry and EFI is exotic.
> > >
> > > > I have also been thru your installer docs
> > >
> > > Mine ? I only write docs about things like optical drives or ISO
> > > 9660 filesystems.
> > > By the latter i happen to be involved in the first boot step of
> > > most Debian ISOs.
> > >
> > > > I feel like I'm playing pin the tale on the donkey, blindfolded.
> > >
> > > As said, i think that David and didier.gaumet show more clue than
> > > i do. So try to get a donkey with tail already attached.
> > >
> > > I understand that David's advise is to look at
> > >   https://raspbian.org/RaspbianImages
> > >   https://www.raspberrypi.org/downloads/raspbian/
> > >
> > >
> > > Have a nice day :)
> > >
> > > Thomas
> >
> > I'll have to take your word for it.
> >
> > I now have a 5 amp. 5.09 volt supply, hot on gpio pin 2, common on
> > gpio pin 6, and I just rewrote the debian-10.0.0-armhf-netinst.iso
> > to /dev/sdf, took it to the rpi4b, and got that same single, 10ms
> > maybe, flash of the green led.  Next I try this .img,
> > 2019-06-20-raspbian-buster.img
> >
> > And I restarted the ssh daemon on the rpi3, running stretch, so
> > maybe I can log into it and get something done yet today. gah, damn
> > raspian, passwds no good.
> >
> > Change that, reboot after enabling ssh. fixed pw, loggged in.
> >
> > And I |think| the raspian is booting. After quite some activity of
> > the green led, its lit solid now, but no damned video. Looks like
> > the new monitor 

Re: Lost sound after update to 9:9.10

[Sven Joachim]

On 2019-09-08 18:13 +, Juan R. de Silva wrote:

> One of my computers lost sound (both speakers and headphones) after
> update to 9:9.10. User is in "audio" group. No sound neither for the user
> nor for root. Hardware tested with Live Ubuntu 16.04 is alive and
> functioning.
>
> Any suggestions, please?

Try booting with the previous kernel, linux-image-4.9.0-9-amd64.  If
that brings the sound back, file a bug against src:linux (preferably
with reportbug).

Cheers,
   Sven

Re: [HS] Richard stallman chez microsoft !

[ajh-valmer]

J'avais oublié ceci :

Microsoft n'a pas digéré d'avoir perdu la bataille des serveurs,
(ils sont majoritairement sous Linux maintenant),
ainsi que celle des smartphone sous Androïd.

Alors, la firme adopte une nouvelle stratégie, totalement opposée
à celle de S. Ballmer, on "achète" Linux en s'introduisant dans
la maison du logiciel opensource afin de récupérer leurs codes sources
et les utiliser petit à petit pour reprendre le marché (Linux Azure est un 
début).
Les produits opensource de microsoft ne sont pas Libres,
bien entendu, avec une licence trafiquée.

Aux salons Opensummmit (POSS) depuis un bout de temps, 
microsoft s'auto-déclare le numéro UN de l'opensource,
en étant le plus gros sponsor du salon.

Pour terrasser son adversaire, la bonne technique est de l'acheter.
On ne peut pas faire gand chose face à ces géants du GAFAM,
ils sont trop riches.

AJH Valmer

On Sunday 08 September 2019 19:46:30 ajh-valmer wrote:
> On Saturday 07 September 2019 10:19:48 Belaïd wrote:
> > Voilà, avec sa stratégie (soi-disant) d'ouverture, Microsoft grignote
> > encore un peu plus le terrain ...
> 
www.zdnet.fr/blogs/l-esprit-libre/richard-stallman-chez-microsoft-non-on-ne-reve-pas-39890151.htm
> 
> Merci de la nouvelle.
> Coment l'as tu détectée ?
> On dirait presque un poisson d'avril bien avant ou bien après l'heure,
> voire un fake-news.
> Qu'est-ce qui a motivé RMS d'aller tenir une conférence chez microsoft ?
> (l'article dit "qu'il a été invité").
> Et qu'a t-il dit ? (l'article ne le dit pas).
> Aucun autre article, semble t-il, sur cette nouvelle surprenante,
> que le site zdnet.
> L'ennemi juré de RMS a toujours été microsoft.
> Ainsi, difficile de donner un point de vue, il manque trop d'infos.

Lost sound after update to 9:9.10

[Juan R. de Silva]

One of my computers lost sound (both speakers and headphones) after 
update to 9:9.10. User is in "audio" group. No sound neither for the user 
nor for root. Hardware tested with Live Ubuntu 16.04 is alive and 
functioning.

Run bellow without any help:

root@desktop:~# alsactl init
Found hardware: "HDA-Intel" "Realtek ALC892" 
"HDA:10ec0892,10438410,00100302" "0x1043" "0x8410"
Hardware is initialized using a generic method


lspci -v related output bellow:
---
00:1b.0 Audio device: Intel Corporation 6 Series/C200 Series Chipset 
Family High Definition Audio Controller (rev 05)
Subsystem: ASUSTeK Computer Inc. 6 Series/C200 Series Chipset 
Family High Definition Audio Controller
Flags: bus master, fast devsel, latency 0, IRQ 46
Memory at fe70 (64-bit, non-prefetchable) [size=16K]
Capabilities: [50] Power Management version 2
Capabilities: [60] MSI: Enable+ Count=1/1 Maskable- 64bit+
Capabilities: [70] Express Root Complex Integrated Endpoint, MSI 
00
Capabilities: [100] Virtual Channel
Capabilities: [130] Root Complex Link
Kernel driver in use: snd_hda_intel
Kernel modules: snd_hda_intel

01:00.1 Audio device: Advanced Micro Devices, Inc. [AMD/ATI] Redwood HDMI 
Audio [Radeon HD 5000 Series]
Subsystem: PC Partner Limited / Sapphire Technology Redwood HDMI 
Audio [Radeon HD 5000 Series]
Flags: bus master, fast devsel, latency 0, IRQ 47
Memory at fe64 (64-bit, non-prefetchable) [size=16K]
Capabilities: [50] Power Management version 3
Capabilities: [58] Express Legacy Endpoint, MSI 00
Capabilities: [a0] MSI: Enable+ Count=1/1 Maskable- 64bit+
Capabilities: [100] Vendor Specific Information: ID=0001 Rev=1 
Len=010 
Capabilities: [150] Advanced Error Reporting
Kernel driver in use: snd_hda_intel
Kernel modules: snd_hda_intel

inxi -Fx related output bellow:
---
System:Host: desktop Kernel: 4.9.0-11-amd64 x86_64 (64 bit gcc: 6.3.0)
   Desktop: Gnome 3.22.3 (Gtk 3.22.11-1) Distro: Debian GNU/Linux 
9 (stretch)
Machine:   Device: desktop Mobo: ASUSTeK model: P8P67 v: Rev 1.xx
   BIOS: American Megatrends v: 1502 date: 03/02/2011

Audio: Card-1 Advanced Micro Devices [AMD/ATI] Redwood HDMI Audio 
[Radeon HD 5000 Series]
   driver: snd_hda_intel bus-ID: 01:00.1
   Card-2 Intel 6 Series/C200 Series Family High Definition Audio 
Controller
   driver: snd_hda_intel bus-ID: 00:1b.0
   Sound: Advanced Linux Sound Architecture v: k4.9.0-11-amd64

Any suggestions, please?

Thanks

Re: openssl headers missing while installing psiphon on Debian 10

[Tapas Mishra]

On Sun, Sep 8, 2019 at 2:56 PM Ulf Volmer  wrote:
>
> On 08.09.19 01:28, Tapas Mishra wrote:
>
> > configure: error: *** OpenSSL headers missing
>
> Install libssl-dev.

Ok thanks I did install libssl-dev
sudo apt-get install libssl-dev
https://pastebin.com/EYpswMRv
I got following
configure: error: Your OpenSSL headers do not match your
library. Check config.log for details.
If you are sure your installation is consistent, you can disable the check
by running "./configure --without-openssl-header-check".
Full message here https://pastebin.com/pyHiyG3j

I  executed ./configure --without-openssl-header-check
full messages https://pastebin.com/UsEZcgKi
now as the instructions here say https://github.com/thispc/psiphon
to do a make after ./configure

I get a following message

authfd.c: In function ‘ssh_encode_identity_ssh2’:
authfd.c:499:33: error: dereferencing pointer to incomplete type ‘DSA’
{aka ‘struct dsa_st’}
   buffer_put_bignum2(b, key->dsa->p);
 ^~
make: *** [Makefile:129: authfd.o] Error 1
full message of make https://pastebin.com/0eSzXLxm
what should I do now how to proceed from here.
I am on Debian 10.Trying to install psiphon.

All steps that I did combined in one paste with all messages
https://pastebin.com/CuMRuExj

How can I proceed from here?
--
Thanks
Tapas

Re: [HS] Richard stallman chez microsoft !

[ajh-valmer]

On Saturday 07 September 2019 10:19:48 Belaïd wrote:
> Voilà, avec sa stratégie (soi-disant) d'ouverture, Microsoft grignote
> encore un peu plus le terrain ...
> 
www.zdnet.fr/blogs/l-esprit-libre/richard-stallman-chez-microsoft-non-on-ne-reve-pas-39890151.htm

Merci de la nouvelle.
Coment l'as tu détectée ?

On dirait presque un poisson d'avril bien avant ou bien après l'heure,
voire un fake-news.

Qu'est-ce qui a motivé RMS d'aller tenir une conférence chez microsoft ?
(l'article dit "qu'il a été invité").

Et qu'a t-il dit ? (l'article ne le dit pas).

Aucun autre article, semble t-il, sur cette nouvelle surprenante,
que le site zdnet.

L'ennemi juré de RMS a toujours été microsoft.

Ainsi, difficile de donner un point de vue, il manque trop d'infos.

Bonne soirée à tous,

A. Valmer

Re: attempted install of buster arm64 net-install on rp4 fails instantly

[Gene Heskett]

On Sunday 08 September 2019 11:34:20 Gene Heskett wrote:

> On Sunday 08 September 2019 08:35:51 Thomas Schmitt wrote:
> > Hi,
> >
> > i now realize that Gene by "rp4" indicated some known Raspberry
> > system. So my proposal about netboot might be hopeless according to
> > the answers of David and didier.gaumet.
> >
> > 
> >-- But there is technical stuff left to discuss:
> >
> > I wrote:
> > > > SD card seems to be the intended target for netboot images.
> > > > http://ftp.debian.org/debian/dists/stable/main/installer-arm64/c
> > > >ur rent/images/netboot/SD-card-images/ [...]
> > > > Those are not ISO9660 but rather partitioned images with a FAT32
> > > > filesystem:
> > > > [...]
> > > >   Device  Boot StartEnd Sectors  Size Id
> > > > Type firmware.a64-olinuxino.img1 * 2048 19  197952 96.7M
> > > > c W95
> >
> > Gene Heskett wrote:
> > > What does this do that the iso doesn't,
> >
> > Those are images for disk-like devices only and they don't look like
> > they are supposed to boot directly via EFI. The Debian arm64 ISOs on
> > the other hand offer typical EFI boot equipment.
> >
> > > and note it takes a windows  machine to follow those instructions.
> >
> > If you mean
> >
> > http://ftp.debian.org/debian/dists/stable/main/installer-arm64/curre
> >nt /images/netboot/SD-card-images/README.concatenateable_images then
> > immediately before the MS-Windows instructions, i read:
> >
> >   "To create a complete image from the two parts on Linux systems,
> > you can use zcat as follows:
> >  zcat firmware..img.gz partition.img.gz >
> > complete_image.img "
> > I inspected both image parts by gunzip-ing them:
> >
> >   $ gunzip firmware.a64-olinuxino.img.gz
> >   $ /sbin/fdisk -lu firmware.a64-olinuxino.img
> >   ... lots of message lines ...
> >   $ gunzip partition.img.gz
> >   $ sudo mount partition.img /mnt/fat
> >   $ find /mnt/fat | less
> >
> > So i would simply provide another useless use of cat with my
> > superuser hat on:
> >
> >   # cat firmware.a64-olinuxino.img partition.img | dd bs=4096
> > of=/dev/sdf1
> >
> > (Insert "sudo" where needed, if your system has no superuser.)
> >
> > > I think what I will do next is send gparted to create a gpt table
> > > with a fat32 first partition of a gig or so, and do an ext4 on the
> > > rest of it. Then put the iso on sdf.
> >
> > The ISO brings its own partition table. Your gparted work will be
> > wasted. See the second grey box at
> >   https://wiki.debian.org/RepackBootableISO#arm64_release_9.4.0
> > which shows fdisk output for arm64 ISOs.
> > (Nothing did change in the ISO partitioning for 10.0 since 9.4.)
> >
> > > Since these come as NTFS formatted cards these days,
> >
> > Filesystems on the card get overwritten by the ISO or at least lose
> > their entry in the partition table, if they are not reached by dd's
> > work.
> >
> > > I would think that
> > > writing the iso to /dev/sdf would at least start the install, and
> > > the installers disk utils could take care of the rest.  What the
> > > iso wrote should be all that counts.
> >
> > As said, Debian arm64 netinst ISOs offer boot entries for EFI.
> > Googling "raspberry rp4 efi" shows that the combination of Raspberry
> > and EFI is exotic.
> >
> > > I have also been thru your installer docs
> >
> > Mine ? I only write docs about things like optical drives or ISO
> > 9660 filesystems.
> > By the latter i happen to be involved in the first boot step of most
> > Debian ISOs.
> >
> > > I feel like I'm playing pin the tale on the donkey, blindfolded.
> >
> > As said, i think that David and didier.gaumet show more clue than i
> > do. So try to get a donkey with tail already attached.
> >
> > I understand that David's advise is to look at
> >   https://raspbian.org/RaspbianImages
> >   https://www.raspberrypi.org/downloads/raspbian/
> >
> >
> > Have a nice day :)
> >
> > Thomas
>
> I'll have to take your word for it.
>
> I now have a 5 amp. 5.09 volt supply, hot on gpio pin 2, common on
> gpio pin 6, and I just rewrote the debian-10.0.0-armhf-netinst.iso to
> /dev/sdf, took it to the rpi4b, and got that same single, 10ms maybe,
> flash of the green led.  Next I try this .img,
> 2019-06-20-raspbian-buster.img
>
> And I restarted the ssh daemon on the rpi3, running stretch, so maybe
> I can log into it and get something done yet today. gah, damn raspian,
> passwds no good.
>
> Change that, reboot after enabling ssh. fixed pw, loggged in.
>
> And I |think| the raspian is booting. After quite some activity of the
> green led, its lit solid now, but no damned video. Looks like the new
> monitor I bought isn't working, its stuck on the vga input, and the
> menu button to change it isn't working, so back to wallies with $80
> worth of junk.
>
> Later.
>
> Cheers, Gene Heskett

Took it back and got another, same exact results, monitor says no input, 
menu button does not work so I can't select/test the 

Re: attempted install of buster arm64 net-install on rp4 fails instantly

[Gene Heskett]

On Sunday 08 September 2019 12:49:14 Curt wrote:

> On 2019-09-08, Gene Heskett  wrote:
> >> > What does this do that the iso doesn't, and note it takes a
> >> > windows machine to follow those instructions.
> >>
> >> Eh? How do you work that out?
> >
> > I won't, theres always another way to skin that cat. ;-) Time, if I
> > have it, will sort that out.
>
> He means how did you derive that erroneous notion from the
> 'README.concatenateable_images,' which says:
>
>  To create a complete image from the two parts on Linux systems, you
> can use zcat as follows:
>
>zcat firmware..img.gz partition.img.gz >
> complete_image.img

zcat is not a cmnd I use, so I didn't recognize it.  Besides, it more of 
the zero info names that one isn't going to remember a week down the log 
when one trips over partition.img.gz and wonders why it was that I 
downloaded it.

So my bad I guess.
> (In fact, in Windows the process is more involved.)
>
> > Cheers, Gene Heskett


Cheers, Gene Heskett
-- 
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
If we desire respect for the law, we must first make the law respectable.
 - Louis D. Brandeis
Genes Web page 

Re: attempted install of buster arm64 net-install on rp4 fails instantly

[tomas]

On Sun, Sep 08, 2019 at 07:01:00PM +0200, deloptes wrote:
> to...@tuxteam.de wrote:

[...]

> > I've the hunch that the firmware also has to be on the top-level dir,
> > but haven't checked yet.
> > 
> > Note that this is all Rpi 3B+
> 
> yes kernel has to be there in the root of the partition

Thanks for confirming :)

Cheers
-- t


signature.asc
Description: Digital signature

Re: attempted install of buster arm64 net-install on rp4 fails instantly

[deloptes]

to...@tuxteam.de wrote:

> AFAIK it also had to be the *first* partition on the medium. I don't know
> whether the kernel has to be on that same partition (but that's the only
> configuration I tested).
> 
> I've the hunch that the firmware also has to be on the top-level dir,
> but haven't checked yet.
> 
> Note that this is all Rpi 3B+

yes kernel has to be there in the root of the partition

Re: attempted install of buster arm64 net-install on rp4 fails instantly

[Curt]

On 2019-09-08, Gene Heskett  wrote:
>> >
>> > What does this do that the iso doesn't, and note it takes a windows
>> > machine to follow those instructions.
>>
>> Eh? How do you work that out?
>
> I won't, theres always another way to skin that cat. ;-) Time, if I have 
> it, will sort that out.

He means how did you derive that erroneous notion from the
'README.concatenateable_images,' which says:
 
 To create a complete image from the two parts on Linux systems, you can
 use zcat as follows:

   zcat firmware..img.gz partition.img.gz > complete_image.img

(In fact, in Windows the process is more involved.)

> Cheers, Gene Heskett


-- 
“We are all in the gutter, but some of us are looking at the stars.” 
― Oscar Wilde, Lady Windermere's Fan

Re: missing /sys/devices/system/cpu/cpu0/cpufreq/scaling_cur_freq on rasbperry pi 3b

[Reco]

Hi.

On Sun, Sep 08, 2019 at 05:39:36PM +0200, basti wrote:
> Hello,
> has someone run debian on rpi 3b or 3b+?
> can anybody confirm this
> https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=939284 bug?
> kernel  4.19.0-6-arm64 does not solve this problem.

Confirming. And it's not going to change unless someone ports
bcm2835-cpufreq.c [1] to the mainline kernel.

Reco

[1] 
https://github.com/raspberrypi/linux/blob/rpi-4.1.y/drivers/cpufreq/bcm2835-cpufreq.c

missing /sys/devices/system/cpu/cpu0/cpufreq/scaling_cur_freq on rasbperry pi 3b

[basti]

Hello,
has someone run debian on rpi 3b or 3b+?
can anybody confirm this
https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=939284 bug?
kernel  4.19.0-6-arm64 does not solve this problem.

Best Regards

Re: attempted install of buster arm64 net-install on rp4 fails instantly

[Gene Heskett]

On Sunday 08 September 2019 10:48:22 deloptes wrote:

> Gene Heskett wrote:
> > And that contained the seeds of my boot problem, the ntfs cards need
> > to have a fat32 re-formatting. NTFS/SDXC doesn't work.
>
> No idea how the 4b works but the older 2b needed vfat because of the
> firmware. Might be still the same. Your idea to create one smaller
> partition with vfat was good. This is what I intend to do as soon as I
> get the 4b, but now I wonder if I would need it at all. If netboot
> works, then it would be only the TFTP server

I would not figure on the netboot working before bullseye. ;-)

Cheers, Gene Heskett
-- 
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
If we desire respect for the law, we must first make the law respectable.
 - Louis D. Brandeis
Genes Web page 

Re: attempted install of buster arm64 net-install on rp4 fails instantly

[Gene Heskett]

On Sunday 08 September 2019 08:35:51 Thomas Schmitt wrote:

> Hi,
>
> i now realize that Gene by "rp4" indicated some known Raspberry
> system. So my proposal about netboot might be hopeless according to
> the answers of David and didier.gaumet.
>
> --
> But there is technical stuff left to discuss:
>
> I wrote:
> > > SD card seems to be the intended target for netboot images.
> > > http://ftp.debian.org/debian/dists/stable/main/installer-arm64/cur
> > >rent/images/netboot/SD-card-images/ [...]
> > > Those are not ISO9660 but rather partitioned images with a FAT32
> > > filesystem:
> > > [...]
> > >   Device  Boot StartEnd Sectors  Size Id
> > > Type firmware.a64-olinuxino.img1 * 2048 19  197952 96.7M 
> > > c W95
>
> Gene Heskett wrote:
> > What does this do that the iso doesn't,
>
> Those are images for disk-like devices only and they don't look like
> they are supposed to boot directly via EFI. The Debian arm64 ISOs on
> the other hand offer typical EFI boot equipment.
>
> > and note it takes a windows  machine to follow those instructions.
>
> If you mean
>  
> http://ftp.debian.org/debian/dists/stable/main/installer-arm64/current
>/images/netboot/SD-card-images/README.concatenateable_images then
> immediately before the MS-Windows instructions, i read:
>
>   "To create a complete image from the two parts on Linux systems, you
> can use zcat as follows:
>  zcat firmware..img.gz partition.img.gz >
> complete_image.img "
> I inspected both image parts by gunzip-ing them:
>
>   $ gunzip firmware.a64-olinuxino.img.gz
>   $ /sbin/fdisk -lu firmware.a64-olinuxino.img
>   ... lots of message lines ...
>   $ gunzip partition.img.gz
>   $ sudo mount partition.img /mnt/fat
>   $ find /mnt/fat | less
>
> So i would simply provide another useless use of cat with my superuser
> hat on:
>
>   # cat firmware.a64-olinuxino.img partition.img | dd bs=4096
> of=/dev/sdf1
>
> (Insert "sudo" where needed, if your system has no superuser.)
>
> > I think what I will do next is send gparted to create a gpt table
> > with a fat32 first partition of a gig or so, and do an ext4 on the
> > rest of it. Then put the iso on sdf.
>
> The ISO brings its own partition table. Your gparted work will be
> wasted. See the second grey box at
>   https://wiki.debian.org/RepackBootableISO#arm64_release_9.4.0
> which shows fdisk output for arm64 ISOs.
> (Nothing did change in the ISO partitioning for 10.0 since 9.4.)
>
> > Since these come as NTFS formatted cards these days,
>
> Filesystems on the card get overwritten by the ISO or at least lose
> their entry in the partition table, if they are not reached by dd's
> work.
>
> > I would think that
> > writing the iso to /dev/sdf would at least start the install, and
> > the installers disk utils could take care of the rest.  What the iso
> > wrote should be all that counts.
>
> As said, Debian arm64 netinst ISOs offer boot entries for EFI.
> Googling "raspberry rp4 efi" shows that the combination of Raspberry
> and EFI is exotic.
>
> > I have also been thru your installer docs
>
> Mine ? I only write docs about things like optical drives or ISO 9660
> filesystems.
> By the latter i happen to be involved in the first boot step of most
> Debian ISOs.
>
> > I feel like I'm playing pin the tale on the donkey, blindfolded.
>
> As said, i think that David and didier.gaumet show more clue than i
> do. So try to get a donkey with tail already attached.
>
> I understand that David's advise is to look at
>   https://raspbian.org/RaspbianImages
>   https://www.raspberrypi.org/downloads/raspbian/
>
>
> Have a nice day :)
>
> Thomas
I'll have to take your word for it.

I now have a 5 amp. 5.09 volt supply, hot on gpio pin 2, common on gpio 
pin 6, and I just rewrote the debian-10.0.0-armhf-netinst.iso 
to /dev/sdf, took it to the rpi4b, and got that same single, 10ms maybe, 
flash of the green led.  Next I try this .img, 
2019-06-20-raspbian-buster.img

And I restarted the ssh daemon on the rpi3, running stretch, so maybe I 
can log into it and get something done yet today. gah, damn raspian, 
passwds no good.

Change that, reboot after enabling ssh. fixed pw, loggged in.

And I |think| the raspian is booting. After quite some activity of the 
green led, its lit solid now, but no damned video. Looks like the new 
monitor I bought isn't working, its stuck on the vga input, and the menu 
button to change it isn't working, so back to wallies with $80 worth of 
junk.

Later.

Cheers, Gene Heskett
-- 
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
If we desire respect for the law, we must first make the law respectable.
 - Louis D. Brandeis
Genes Web page 

Re: attempted install of buster arm64 net-install on rp4 fails instantly

[tomas]

On Sun, Sep 08, 2019 at 04:48:22PM +0200, deloptes wrote:
> Gene Heskett wrote:
> 
> > And that contained the seeds of my boot problem, the ntfs cards need to
> > have a fat32 re-formatting. NTFS/SDXC doesn't work.
> 
> No idea how the 4b works but the older 2b needed vfat because of the
> firmware [...]

AFAIK it also had to be the *first* partition on the medium. I don't know
whether the kernel has to be on that same partition (but that's the only
configuration I tested).

I've the hunch that the firmware also has to be on the top-level dir,
but haven't checked yet.

Note that this is all Rpi 3B+

Cheers
-- t


signature.asc
Description: Digital signature

Re: attempted install of buster arm64 net-install on rp4 fails instantly

[deloptes]

Thomas Schmitt wrote:

> As said, Debian arm64 netinst ISOs offer boot entries for EFI.
> Googling "raspberry rp4 efi" shows that the combination of Raspberry and
> EFI is exotic.

Thomas,
the RPI had its own boot process - you can imaging booting mobile phone. Not
sure what they did for the RPI4b now - at least it looks like the
bootloader now is on the EEPROM - it said in the GPU AFAIR.
the boot partition must be vfat in the beginning of the disk, so that the
firmware can load the bootloader. It might be not the case now with the
RPI4

Re: attempted install of buster arm64 net-install on rp4 fails instantly

[deloptes]

Gene Heskett wrote:

> And that contained the seeds of my boot problem, the ntfs cards need to
> have a fat32 re-formatting. NTFS/SDXC doesn't work.

No idea how the 4b works but the older 2b needed vfat because of the
firmware. Might be still the same. Your idea to create one smaller
partition with vfat was good. This is what I intend to do as soon as I get
the 4b, but now I wonder if I would need it at all. If netboot works, then
it would be only the TFTP server

Re: attempted install of buster arm64 net-install on rp4 fails instantly

[Gene Heskett]

On Sunday 08 September 2019 08:32:30 deloptes wrote:

> Gene Heskett wrote:
> > I have also been thru your installer docs, without finding specific
> > procedures for initializing the u-sd cards its being installed from.
> > Did I miss that link in my searches?  Without that I feel like I'm
> > playing pin the tale on the donkey, blindfolded.
>
> I think you should start reading here
> https://www.raspberrypi.org/documentation/installation/

And that contained the seeds of my boot problem, the ntfs cards need to 
have a fat32 re-formatting. NTFS/SDXC doesn't work.

Cheers, Gene Heskett
-- 
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
If we desire respect for the law, we must first make the law respectable.
 - Louis D. Brandeis
Genes Web page 

Re: attempted install of buster arm64 net-install on rp4 fails instantly

[Gene Heskett]

On Sunday 08 September 2019 07:42:36 Michael Howard wrote:

> On 08/09/2019 12:39, Gene Heskett wrote:
> > On Sunday 08 September 2019 03:59:01 Thomas Schmitt wrote:
> >> Hi,
> >>
> >> Gene Heskett wrote:
> >>> debian-10.0.0-armhf-netinst.iso
> >>> debian-10.0.0-arm64-netinst.iso
> >>
> >> SD card seems to be the intended target for netboot images.
> >> E.g.
> >>
> >> http://ftp.debian.org/debian/dists/stable/main/installer-arm64/curr
> >>ent /images/netboot/SD-card-images/
> >>
> >> Maybe worth a try, if you can get advise about the appropriate
> >> image for your ARM machine.
> >>
> >> Those are not ISO9660 but rather partitioned images with a FAT32
> >> filesystem:
> >>
> >>$ /sbin/fdisk -lu firmware.a64-olinuxino.img
> >>...
> >>Disklabel type: dos
> >>...
> >>Device  Boot StartEnd Sectors  Size Id
> >> Type firmware.a64-olinuxino.img1 * 2048 19  197952 96.7M  c
> >> W95 FAT32 (LBA)
> >>
> >>
> >> Have a nice day :)
> >>
> >> Thomas
> >
> > What does this do that the iso doesn't, and note it takes a windows
> > machine to follow those instructions.
>
> Eh? How do you work that out?

I won't, theres always another way to skin that cat. ;-) Time, if I have 
it, will sort that out.

Cheers, Gene Heskett
-- 
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
If we desire respect for the law, we must first make the law respectable.
 - Louis D. Brandeis
Genes Web page 

Re: attempted install of buster arm64 net-install on rp4 fails instantly

[Gene Heskett]

On Sunday 08 September 2019 06:12:08 David wrote:

> On Sun, 8 Sep 2019 at 19:51, David  wrote:
> > On Sun, 8 Sep 2019 at 06:59, Gene Heskett  
wrote:
> > > [...] What did I do wrong [...] ?
> >
> > I don't know about the Debian ARM images, but if
> > they were relevant then I assume they would be
> > mentioned at the links above.
>
> And here's a very recent discussion of what is required to
> install Ubuntu server in either 32 or 64bit version on
> Raspberry Pi 4, which may be helpful:
> https://jamesachambers.com/raspberry-pi-ubuntu-server-18-04-2-installa
>tion-guide/
>
> I don't have any special needs for my Pi 4 so it is currently
> running with a Raspbian image, like most people do.
> https://raspbian.org/RaspbianImages
> https://www.raspberrypi.org/downloads/raspbian/

I get the impression its going to take till bullseye to begin to get it 
all in "one sock" for the pi4b. The above is fraught with all sorts of 
patches & fixes, none of which have descriptive enough names to their 
downloads to remember what the hell they do next week.  None of that is 
your (debian's) fault, but it does take time to merge all that into an 
it just works package.

Based on that, I may, probably will, revert to a stretch install on the 
pi3b, and put up with the video frame rate you can count on your 
fingers. It does run the machine fairly well. But I'll also have the 
pi4b setup on the table, and keep it on testing I think, just to watch 
progress.

Unless the RealtimePi guy makes a buster kit for the rpi3b that uses the 
new video libraries, that would be awful tempting to play with on a 
pi3b. So far, his realtime kernels are too old to use the new video, 
which is nearly 20x faster.

Whoever gets there first, with a stable os is what it boils down to. At 
this point, it looks like raspians 2nd buster release, if the hm2_rpspi 
driver works with it.  Thats a big if. And it must work else I'll be out 
around $200 for all new interfacing hardware.

Thanks for the heads up David, you are going to save me time. Time I may 
not have, but so far he's not doing anything final. I appreciate it, a 
lot.

Cheers, Gene Heskett
-- 
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
If we desire respect for the law, we must first make the law respectable.
 - Louis D. Brandeis
Genes Web page 

Re: attempted install of buster arm64 net-install on rp4 fails instantly

[John Hasler]

Sorry.  I don't know why it went out like that.
-- 
John Hasler 
jhas...@newsguy.com
Elmwood, WI USA

Re: attempted install of buster arm64 net-install on rp4 fails instantly

[John Hasler]

Gene writes:
> I have also been thru your installer docs, without finding specific
> procedures for initializing the u-sd cards its being installed from.

You dd the image directly to the raw device.  What was previously on the
card is irrelevant: it gets overwritten.

I recently installed Armbian on an Espressobin (uses a Marvell chip).
These installs can be a bit gnarly.  I had to upgrade the Uboot loader
and patch some variables on the chip.
-- 
John Hasler 
jhas...@newsguy.com
Elmwood, WI USA

Re: attempted install of buster arm64 net-install on rp4 fails instantly

[Thomas Schmitt]

Hi,

i now realize that Gene by "rp4" indicated some known Raspberry system.
So my proposal about netboot might be hopeless according to the answers
of David and didier.gaumet.

--
But there is technical stuff left to discuss:

I wrote:
> > SD card seems to be the intended target for netboot images.
> > http://ftp.debian.org/debian/dists/stable/main/installer-arm64/current/images/netboot/SD-card-images/
> > [...]
> > Those are not ISO9660 but rather partitioned images with a FAT32
> > filesystem:
> > [...]
> >   Device  Boot StartEnd Sectors  Size Id Type
> >   firmware.a64-olinuxino.img1 * 2048 19  197952 96.7M  c W95

Gene Heskett wrote:
> What does this do that the iso doesn't,

Those are images for disk-like devices only and they don't look like
they are supposed to boot directly via EFI. The Debian arm64 ISOs on the
other hand offer typical EFI boot equipment.


> and note it takes a windows  machine to follow those instructions.

If you mean
  
http://ftp.debian.org/debian/dists/stable/main/installer-arm64/current/images/netboot/SD-card-images/README.concatenateable_images
then immediately before the MS-Windows instructions, i read:

  "To create a complete image from the two parts on Linux systems, you can
   use zcat as follows:
 zcat firmware..img.gz partition.img.gz > complete_image.img
   "
I inspected both image parts by gunzip-ing them:

  $ gunzip firmware.a64-olinuxino.img.gz
  $ /sbin/fdisk -lu firmware.a64-olinuxino.img
  ... lots of message lines ...
  $ gunzip partition.img.gz
  $ sudo mount partition.img /mnt/fat
  $ find /mnt/fat | less

So i would simply provide another useless use of cat with my superuser hat
on:

  # cat firmware.a64-olinuxino.img partition.img | dd bs=4096 of=/dev/sdf1

(Insert "sudo" where needed, if your system has no superuser.)


> I think what I will do next is send gparted to create a gpt table with a
> fat32 first partition of a gig or so, and do an ext4 on the rest of it.
> Then put the iso on sdf.

The ISO brings its own partition table. Your gparted work will be wasted.
See the second grey box at
  https://wiki.debian.org/RepackBootableISO#arm64_release_9.4.0
which shows fdisk output for arm64 ISOs.
(Nothing did change in the ISO partitioning for 10.0 since 9.4.)


> Since these come as NTFS formatted cards these days,

Filesystems on the card get overwritten by the ISO or at least lose their
entry in the partition table, if they are not reached by dd's work.


> I would think that
> writing the iso to /dev/sdf would at least start the install, and the
> installers disk utils could take care of the rest.  What the iso wrote
> should be all that counts.

As said, Debian arm64 netinst ISOs offer boot entries for EFI.
Googling "raspberry rp4 efi" shows that the combination of Raspberry and
EFI is exotic.


> I have also been thru your installer docs

Mine ? I only write docs about things like optical drives or ISO 9660
filesystems.
By the latter i happen to be involved in the first boot step of most
Debian ISOs.


> I feel like I'm playing pin the tale on the donkey, blindfolded.

As said, i think that David and didier.gaumet show more clue than i do.
So try to get a donkey with tail already attached.

I understand that David's advise is to look at
  https://raspbian.org/RaspbianImages
  https://www.raspberrypi.org/downloads/raspbian/


Have a nice day :)

Thomas

Re: attempted install of buster arm64 net-install on rp4 fails instantly

[deloptes]

Gene Heskett wrote:

> I have also been thru your installer docs, without finding specific
> procedures for initializing the u-sd cards its being installed from. Did
> I miss that link in my searches?  Without that I feel like I'm playing
> pin the tale on the donkey, blindfolded.

I think you should start reading here
https://www.raspberrypi.org/documentation/installation/

Re: attempted install of buster arm64 net-install on rp4 fails instantly

[Gene Heskett]

On Sunday 08 September 2019 06:05:37 didier.gau...@gmail.com wrote:

> Debian Buster is not yet ready for Raspberry Pi 4:
>
> https://wiki.debian.org/RaspberryPiImages
> https://www.raspberrypi.org/documentation/installation/

Potential timeline?

Cheers, Gene Heskett
-- 
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
If we desire respect for the law, we must first make the law respectable.
 - Louis D. Brandeis
Genes Web page 

Re: attempted install of buster arm64 net-install on rp4 fails instantly

[Gene Heskett]

On Sunday 08 September 2019 05:51:43 David wrote:

> On Sun, 8 Sep 2019 at 06:59, Gene Heskett  wrote:
> > [...] What did I do wrong [...] ?
>
> I have attempted to find relevant info from the Debian project
> for you. I found these two pages, which seem to be recently
> updated by active developers:
> [1] https://salsa.debian.org/raspi-team/image-specs
> [2] https://wiki.debian.org/RaspberryPiImages
>
> Quoting from [2] above:
> "The new (June 2019) Raspberry Pi 4 is not yet supported
> (the kernel does not yet have a Device Tree for it) "
>
I may have, as of yesterday, a testable gpio/spi driver for the pi4b, it 
appears the major change is the base address of the gpio stuffs.  One 
conditional define, maybe.

> I don't know about the Debian ARM images, but if
> they were relevant then I assume they would be
> mentioned at the links above.

I'd assume so too, but then the pi4 announce no doubt caught the debian 
folks busier than that famous cat covering his output on a tin roof.
truly good support if lots of changes may wait for bullseye.  That IMO 
can be laid on the Foundations doorstep, the deeper the better, as they 
should have been sharing that under an NDI of some sort.  Catching your 
folks out does not a pretty picture of the Foundation paint.

Thanks David.

Cheers, Gene Heskett
-- 
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
If we desire respect for the law, we must first make the law respectable.
 - Louis D. Brandeis
Genes Web page 

Re: Re: issues with initramfs after recently updating buster

[ernst doubt]

Thanks so much Pascal Hambourg for all the great info.

  best,
   ~e

-- 
https://edoubt.tumblr.com
https://medium.com/@cderr/

signature.asc
Description: This is a digitally signed message part.

Re: attempted install of buster arm64 net-install on rp4 fails instantly

[Gene Heskett]

On Sunday 08 September 2019 04:03:00 deloptes wrote:

> Gene Heskett wrote:
> > I have a bigger 4 amp supply  but haven't rigged it with output
> > jumpers yet.
>
> But in the specs of the RPI4 it says it requires a 3A PWR supply.

Thats not listed on the multi-language paper in the rpi4 bag. Not your 
(debians fault) but some lawyer wrote that to specificly remove any 
helpfull info, like how to power it up.

I'll get that going with a 4 or 5 amp today. I should have at least one 
of those in the spares box yet.

> However I think it is possible that you have another issue with the
> SD-CARD.
>
> It looks to me that now with RPI4 and the bootloader placed on the
> device.
>
> What I do not understand is if the net boot works or not, cause it is
> saying it will be enabled by optional updates later.

I saw that someplace too, and am confused that net booting was even 
mentioned. Thats a whole new critter.

Net boot is not net installer, by quite a lengthy row of apple trees.

I'll get a new bigger supply hooked up today and let this list sort out 
how to prep a 64Gig, ntfs formatted u-sd card. I believe I now have an 
spi driver that will work with both a pi3b and the new pi4b. Testable at 
least.

Thanks all.

Cheers, Gene Heskett
-- 
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
If we desire respect for the law, we must first make the law respectable.
 - Louis D. Brandeis
Genes Web page 

Re: attempted install of buster arm64 net-install on rp4 fails instantly

[Michael Howard]

On 08/09/2019 12:39, Gene Heskett wrote:

On Sunday 08 September 2019 03:59:01 Thomas Schmitt wrote:


Hi,

Gene Heskett wrote:

debian-10.0.0-armhf-netinst.iso
debian-10.0.0-arm64-netinst.iso

SD card seems to be the intended target for netboot images.
E.g.
  
http://ftp.debian.org/debian/dists/stable/main/installer-arm64/current

/images/netboot/SD-card-images/

Maybe worth a try, if you can get advise about the appropriate image
for your ARM machine.

Those are not ISO9660 but rather partitioned images with a FAT32
filesystem:

   $ /sbin/fdisk -lu firmware.a64-olinuxino.img
   ...
   Disklabel type: dos
   ...
   Device  Boot StartEnd Sectors  Size Id Type
   firmware.a64-olinuxino.img1 * 2048 19  197952 96.7M  c W95
FAT32 (LBA)


Have a nice day :)

Thomas

What does this do that the iso doesn't, and note it takes a windows
machine to follow those instructions.


Eh? How do you work that out?

--
Mike Howard

Re: attempted install of buster arm64 net-install on rp4 fails instantly

[Gene Heskett]

On Sunday 08 September 2019 03:59:01 Thomas Schmitt wrote:

> Hi,
>
> Gene Heskett wrote:
> > debian-10.0.0-armhf-netinst.iso
> > debian-10.0.0-arm64-netinst.iso
>
> SD card seems to be the intended target for netboot images.
> E.g.
>  
> http://ftp.debian.org/debian/dists/stable/main/installer-arm64/current
>/images/netboot/SD-card-images/
>
> Maybe worth a try, if you can get advise about the appropriate image
> for your ARM machine.
>
> Those are not ISO9660 but rather partitioned images with a FAT32
> filesystem:
>
>   $ /sbin/fdisk -lu firmware.a64-olinuxino.img
>   ...
>   Disklabel type: dos
>   ...
>   Device  Boot StartEnd Sectors  Size Id Type
>   firmware.a64-olinuxino.img1 * 2048 19  197952 96.7M  c W95
> FAT32 (LBA)
>
>
> Have a nice day :)
>
> Thomas

What does this do that the iso doesn't, and note it takes a windows 
machine to follow those instructions.  No such critter here, hasn't been
been for 50+ years.

I think what I will do next is send gparted to create a gpt table with a 
fat32 first partition of a gig or so, and do an ext4 on the rest of it. 
Then put the iso on sdf.  By using an sd card that big, I have never 
worn one out unless it was a sandisk. Those seem be garbage from the 
gitgo. I've even had one go read-only on the first write with dd.

Since these come as NTFS formatted cards these days, I would think that 
writing the iso to /dev/sdf would at least start the install, and the 
installers disk utils could take care of the rest.  What the iso wrote 
should be all that counts. 

I had one of busters rc3's (arm64) installed and running but got 
sidetracked because while it ran nice, amanda killed it trying to do a 
backup. I think thats an amanda build problem as it seems to apply to 
any arm64 install as it would also kill an armbian install on a rock64. 
armhf's worked fine. arm64's give BUS err.

Now, from my searches just now, I can't find a 10.1 version of these 
installers.  I thought I read that the mirrors s/b updated by now, but 
the web pages apparently have not been, so the new release is still 
hidden from view.  Or is my google-foo broken again?

I have also been thru your installer docs, without finding specific 
procedures for initializing the u-sd cards its being installed from. Did 
I miss that link in my searches?  Without that I feel like I'm playing 
pin the tale on the donkey, blindfolded.

Thanks Thomas.

Cheers, Gene Heskett
-- 
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
If we desire respect for the law, we must first make the law respectable.
 - Louis D. Brandeis
Genes Web page 

Re: AMD Vega 6 on Debian 10

[Alessandro Vesely]

On Thu 05/Sep/2019 22:00:58 +0200 Étienne Mollier wrote:
> Ale, on 2019-09-05:
>> On Wed 04/Sep/2019 20:12:55 +0200 Étienne Mollier wrote:
>> > Thank you Sven for pointing this out!  You're right, I've been
>> > confused by personal experience with other chips.  Installation
>> > of "firmware-amd-graphics" should be sufficient in that case.
>> 
>> For the record, it doesn't work.  I additionally tinkered for a while with 
>> the
>> drivers supplied by AMD, but didn't succeed.  Eventually, I gave up.  Ubuntu
>> recognizes it out of the box.
> 
> Out of curiosity, which Ubuntu release did the job?


It was Ubuntu 18.04.3 LTS.
http://releases.ubuntu.com/18.04/ubuntu-18.04.3-desktop-amd64.iso


> It could give a clue as of required firmware or drivers levels
> for your integrated card, or any other missing component.


Given that Ubuntu is based on Debian, drawing liberally on their work is not
only shameless, it is a savvy arrangement to avoid doing the same work twice.


Best
Ale



signature.asc
Description: OpenPGP digital signature

Re: attempted install of buster arm64 net-install on rp4 fails instantly

[didier . gaumet]

Debian Buster is not yet ready for Raspberry Pi 4:

https://wiki.debian.org/RaspberryPiImages
https://www.raspberrypi.org/documentation/installation/

Re: attempted install of buster arm64 net-install on rp4 fails instantly

[David]

On Sun, 8 Sep 2019 at 19:51, David  wrote:
> On Sun, 8 Sep 2019 at 06:59, Gene Heskett  wrote:

> > [...] What did I do wrong [...] ?

> I don't know about the Debian ARM images, but if
> they were relevant then I assume they would be
> mentioned at the links above.

And here's a very recent discussion of what is required to
install Ubuntu server in either 32 or 64bit version on
Raspberry Pi 4, which may be helpful:
https://jamesachambers.com/raspberry-pi-ubuntu-server-18-04-2-installation-guide/

I don't have any special needs for my Pi 4 so it is currently
running with a Raspbian image, like most people do.
https://raspbian.org/RaspbianImages
https://www.raspberrypi.org/downloads/raspbian/

Re: attempted install of buster arm64 net-install on rp4 fails instantly

[David]

On Sun, 8 Sep 2019 at 06:59, Gene Heskett  wrote:
>
> [...] What did I do wrong [...] ?

I have attempted to find relevant info from the Debian project
for you. I found these two pages, which seem to be recently
updated by active developers:
[1] https://salsa.debian.org/raspi-team/image-specs
[2] https://wiki.debian.org/RaspberryPiImages

Quoting from [2] above:
"The new (June 2019) Raspberry Pi 4 is not yet supported
(the kernel does not yet have a Device Tree for it) "

I don't know about the Debian ARM images, but if
they were relevant then I assume they would be
mentioned at the links above.

Re: openssl headers missing while installing psiphon on Debian 10

[Ulf Volmer]

On 08.09.19 01:28, Tapas Mishra wrote:

> configure: error: *** OpenSSL headers missing

Install libssl-dev.

Best regards
Ulf

Re: issues with initramfs after recently updating buster

[Pascal Hambourg]

Le 08/09/2019 à 02:28, ernst doubt a écrit :


Today I saw the following errors on one of the machines (an HP EliteBook
laptop) that I upgraded today:

update-initramfs: Generating /boot/initrd.img-4.19.0-5-amd64
cryptsetup: WARNING: The initramfs image may not contain cryptsetup binaries
 nor crypto modules. If that's on purpose, you may want to uninstall the
 'cryptsetup-initramfs' package in order to disable the cryptsetup
initramfs
 integration and avoid this warning.
W: Possible missing firmware /lib/firmware/i915/bxt_dmc_ver1_07.bin for module
i915

(...)

These are warnings, not errors.


So I have two questions. On the HP laptop I do use cryptsetup (because it's
older and I originally ran stretch on that machine -- only my homedir is
encrypted (not the entire LV as is the Dell laptop, which is newer and got a
virgin buster install recently). Is there anything I need to do in order to
make sure that I don't lose support for cryptsetup with a new kernel on the HP
laptop?


No. You need to worry about this warning only if /, /usr (if in a 
separate filesystem) or the swap used for hibernation is encrypted 
because these must be opened by the initramfs. Any other encrypted 
device will be opened later by the main init system.


You can uninstall cryptsetup-initramfs (do not worry if it uninstalls 
cryptsetup too, this is a dummy package) to remove the warning.



My 2nd question is about the missing firmware. How do I get modules i915 and
amdgpu so that my initramfs will be fully functional on these laptops (i'm fine
with installing nonfree packages).


For i915 : install firmware-misc-nonfree.
For amdgpu/radeon : install firmware-amd-graphics.
Or install firwmware-linux-nonfree which will pull both.
All are in the non-free section which must be enabled in sources.list.


On one laptop (the HP), when I try to run 'apt-get upgrade' again, the
following message appears:

The following packages have been kept back:
   linux-headers-amd64 linux-image-amd64

On the other laptop (the Dell), 'apt-get upgrade' produces output that
indicates that only linux-image-amd64 has been kept back.


These are meta-packages which depend on the latest available kernel 
packages version. Upgrading them requires installing new kernel 
packages. But apt-get won't install new packages by default. You can 
upgrade them using "apt-get dist-upgrade" or apt-get --with-new-pkgs 
update" or "apt upgrade" (instead of apt-get). Or just install the 
packages as if they were not installed.

Re: attempted install of buster arm64 net-install on rp4 fails instantly

[deloptes]

Gene Heskett wrote:

> I have a bigger 4 amp supply  but haven't rigged it with output jumpers
> yet.

But in the specs of the RPI4 it says it requires a 3A PWR supply.

However I think it is possible that you have another issue with the SD-CARD. 

It looks to me that now with RPI4 and the bootloader placed on the device.

What I do not understand is if the net boot works or not, cause it is saying
it will be enabled by optional updates later.

regards

Re: attempted install of buster arm64 net-install on rp4 fails instantly

[Thomas Schmitt]

Hi,

Gene Heskett wrote:
> debian-10.0.0-armhf-netinst.iso
> debian-10.0.0-arm64-netinst.iso

SD card seems to be the intended target for netboot images.
E.g.
  
http://ftp.debian.org/debian/dists/stable/main/installer-arm64/current/images/netboot/SD-card-images/

Maybe worth a try, if you can get advise about the appropriate image
for your ARM machine.

Those are not ISO9660 but rather partitioned images with a FAT32 filesystem:

  $ /sbin/fdisk -lu firmware.a64-olinuxino.img
  ...
  Disklabel type: dos
  ...
  Device  Boot StartEnd Sectors  Size Id Type
  firmware.a64-olinuxino.img1 * 2048 19  197952 96.7M  c W95 FAT32 (LBA)


Have a nice day :)

Thomas

Re: attempted install of buster arm64 net-install on rp4 fails instantly

[Gene Heskett]

On Sunday 08 September 2019 02:18:38 David wrote:

> On Sun, 8 Sep 2019 at 06:59, Gene Heskett  wrote:
> > What did I do wrong in dd-ing
> > the net-install iso to the new 64GB u-sd card?
>
> Please give us a link so that we can identify and test whatever
> file you are calling "the net-install iso".
>
> Please show us exactly the "dd" command you used to write
> it to the sd card.
>
> If you do that, it makes it possible that we can duplicate your
> situation and/or reliably answer your question.

i should have changed the subject, thats an armhf iso, the last one I 
tried, but both were tried yesterday, same results.

The md5sum of the arm64 file
root@coyote:pi-buster$ md5sum debian-10.0.0-arm64-netinst.iso
c957472ce923bf6e0afe0f4c6a22d6c2  debian-10.0.0-arm64-netinst.iso


Cheers, Gene Heskett
-- 
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
If we desire respect for the law, we must first make the law respectable.
 - Louis D. Brandeis
Genes Web page 

Re: attempted install of buster arm64 net-install on rp4 fails instantly

[Gene Heskett]

On Sunday 08 September 2019 02:18:38 David wrote:

> On Sun, 8 Sep 2019 at 06:59, Gene Heskett  wrote:
> > What did I do wrong in dd-ing
> > the net-install iso to the new 64GB u-sd card?
>
> Please give us a link so that we can identify and test whatever
> file you are calling "the net-install iso".

debian-10.0.0-armhf-netinst.iso, located on this machines main drive, 
downloaded from a debian site with FF.

md5sum=df55beefd09acd1db38e6e66cb8308c2
>
> Please show us exactly the "dd" command you used to write
> it to the sd card.

sudo dd if=debian-10.0.0-armhf-netinst.iso bs=4096 of=/dev/sdf1

/dev/sdf1 is an unmounted 64GB PNY u-sd card. Original format NTFS.

Mounting it and reading the directory after the dd write looks like I 
think it should, a directory of the iso, not the iso itself.

The rpi4b is being powered from its gpio pins 2=5.11 volts, 6=common from 
a canna kit v2.5 amp supply, which may not be enough amps as theres 2 
keyboard/mouse dongles in the usb2 ports, and a usb3 hub with two SSD's 
plugged into a usb3 (blue port) and a single micro to std hdmi plugged 
into hdmi#1. and a net cable from my garage switch plugged into the cat5 
jack.  No heat sinks, they're someplace between HongKong and WV yet. No 
detectable heating, just the red led is lit, green flashes for a very 
few milliseconds, once.

This same supply ran a rock64 with all that same stuff plugged in w/o a 
problem, running armbian on the rock64.

Dead stable until amanda touched it, crashed. Amanda is the backup 
utility of choice here, has been for 20 years now.
Amanda from your repo's crashes all arm64 installs, but thats a separate, 
different problem we'll solve later. Probably by building it from src, 
I'm used to that. NBD to me.

I have a bigger 4 amp supply  but haven't rigged it with output jumpers 
yet. 
> If you do that, it makes it possible that we can duplicate your
> situation and/or reliably answer your question.


Cheers, Gene Heskett
-- 
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
If we desire respect for the law, we must first make the law respectable.
 - Louis D. Brandeis
Genes Web page 

Re: attempted install of buster arm64 net-install on rp4 fails instantly

[David]

On Sun, 8 Sep 2019 at 06:59, Gene Heskett  wrote:
>
> What did I do wrong in dd-ing
> the net-install iso to the new 64GB u-sd card?

Please give us a link so that we can identify and test whatever
file you are calling "the net-install iso".

Please show us exactly the "dd" command you used to write
it to the sd card.

If you do that, it makes it possible that we can duplicate your
situation and/or reliably answer your question.