Re: Regression in Radeon driver in latest Debian Stable kernel

[Michael Kjörling]

On 17 May 2024 00:12 +0700, from maniku...@gmail.com (Max Nikulin):
> Be realistic, to get the bug fixed, there should be affected persons
> motivated enough to try vanilla kernel or even to build custom kernels with
> provided patches. Developers time is limited and expensive resource. It may
> be directed to fixing other bugs. If you can compare Debian and upstream
> kernels you may get the issue fixed quicker. Direct communication with
> driver developers may be more effective.

Indeed; that's why I suggested trying the vanilla kernels of the same
versions, compiled with the same options, and seeing if the behavior
can be reproduced with those. If the same behavior as seen with the
Debian-packaged kernels can be reproduced with the vanilla kernel,
that would very strongly suggest that whatever this is is either (a)
an upstream issue in the kernel, or (b) not a kernel issue at all.
There's a reason why many upstream maintainers, when faced with a bug
report for a downstream potentially modified version of their code,
will start by saying "try our version".

Adding to the "be realistic", if the issue was an obvious one, it
likely never would have made it into a released kernel at all. So
whatever this is about is unlikely to be obvious. Thus some detective
work is most likely going to be needed; and unless a kernel developer
can reproduce the problem on their own hardware, they'll probably have
to ask you to try things out and report back; whether with debug logs,
more detailed system information, or as Max mentioned building a
kernel with a proposed fix to see what effect, if any, a proposed fix
has on the problem.

Also, especially if you start installing backports kernels as well,
you may want to add a version pin to the kernel you currently have
installed and which does not exhibit the problem to the same degree,
to reduce the risk that it gets purged for being among the older ones
you have installed.

-- 
Michael Kjörling  https://michael.kjorling.se
“Remember when, on the Internet, nobody cared that you were a dog?”

Re: Regression in Radeon driver in latest Debian Stable kernel

[Max Nikulin]

On 16/05/2024 22:53, piorunz wrote:

https://bugzilla.kernel.org/ will they be interested in Debian specific
error? I don't use vanilla kernel so maybe it's Debian only problem.


If you *find* a similar report there then it is likely an upstream 
issue. (Developers of specific driver may use another bug tracker.)


Do you have a recipe how to trigger the bug? Otherwise it may be much 
harder to fix it.


Be realistic, to get the bug fixed, there should be affected persons 
motivated enough to try vanilla kernel or even to build custom kernels 
with provided patches. Developers time is limited and expensive 
resource. It may be directed to fixing other bugs. If you can compare 
Debian and upstream kernels you may get the issue fixed quicker. Direct 
communication with driver developers may be more effective.

Re: Regression in Radeon driver in latest Debian Stable kernel

[piorunz]

On 16/05/2024 12:35, Max Nikulin wrote:

On 16/05/2024 17:35, piorunz wrote:


As much as I would like to try vanilla kernel, I don't want to break my
system. I use Debian Stable, don't know if things would just work with
vanilla kernel.


You may try bookworm-backports kernel 6.6.13+bpo-amd64

You may check https://bugzilla.kernel.org/ for reports similar to your one.


Oh that's great solution, I will try that without risk of breaking my
system, as I always can select previous kernel in the menu, thank you.

https://bugzilla.kernel.org/ will they be interested in Debian specific
error? I don't use vanilla kernel so maybe it's Debian only problem.

--
With kindest regards, Piotr.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org/
⠈⠳⣄

Re: Regression in Radeon driver in latest Debian Stable kernel

[Max Nikulin]

On 16/05/2024 17:35, piorunz wrote:


As much as I would like to try vanilla kernel, I don't want to break my
system. I use Debian Stable, don't know if things would just work with
vanilla kernel.


You may try bookworm-backports kernel 6.6.13+bpo-amd64

You may check https://bugzilla.kernel.org/ for reports similar to your one.

Re: Regression in Radeon driver in latest Debian Stable kernel

[piorunz]

On 15/05/2024 20:55, Michael Kjörling wrote:


You made this bug report less than 48 hours ago. While I can certainly
understand that you would like to see it fixed, that's not an
inordinate amount of time to wait.

What probably _would_ be helpful is to see whether you can recreate
the same scenario with the vanilla kernel.org kernels of the same
versions; confirming that the issue exists with the vanilla 6.1.90
kernel _and_ that the issue goes away when booting the vanilla kernel
of whatever exact version your earlier package is (I'm guessing
6.1.85), both when built with the same configuration options as the
kernels shipped by Debian. That would confirm whether what you are
seeing is something somehow introduced by Debian, or an upstream bug.


Thank you for your reply, I appreciate it.
I found out that this error happened on previous version of the kernel
as well, just found it in the logs. So, it's not an immediate regression
in the latest version of the kernel in Debian. Maybe it's deeper than
that. That lowers the possibility that this is the most recent update
which brought this problem. I sent latest update to the bug report, I
will wait for any reply from developer but no rush, of course.

As much as I would like to try vanilla kernel, I don't want to break my
system. I use Debian Stable, don't know if things would just work with
vanilla kernel.

--
With kindest regards, Piotr.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org/
⠈⠳⣄

Re: Regression in Radeon driver in latest Debian Stable kernel

[Michael Kjörling]

On 15 May 2024 20:40 +0100, from pior...@gmx.com (piorunz):
> I have reported a regression in latest Linux kernel in Debian Stable:
> 
> segfault at amdgpu_dm_atomic_commit_tail
> https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1071080

You made this bug report less than 48 hours ago. While I can certainly
understand that you would like to see it fixed, that's not an
inordinate amount of time to wait.

What probably _would_ be helpful is to see whether you can recreate
the same scenario with the vanilla kernel.org kernels of the same
versions; confirming that the issue exists with the vanilla 6.1.90
kernel _and_ that the issue goes away when booting the vanilla kernel
of whatever exact version your earlier package is (I'm guessing
6.1.85), both when built with the same configuration options as the
kernels shipped by Debian. That would confirm whether what you are
seeing is something somehow introduced by Debian, or an upstream bug.

-- 
Michael Kjörling  https://michael.kjorling.se
“Remember when, on the Internet, nobody cared that you were a dog?”

Regression in Radeon driver in latest Debian Stable kernel

[piorunz]

Hello,

I have reported a regression in latest Linux kernel in Debian Stable:

segfault at amdgpu_dm_atomic_commit_tail
https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1071080

It throws a lot of errors related to AMD GPU every day. I also
experienced full desktop hang, where I had to restart my computer:
https://bugs.kde.org/show_bug.cgi?id=486970

This happens on 6.1.0-21-amd64. Booting previous version 6.1.0-20-amd64
solves all problems.

Can anyone help me with getting attention from some developers? KDE dev
said this is downstream, not KDE problem. Debian bug reported, but no
reply from anyone yet.

--
With kindest regards, Piotr.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org/
⠈⠳⣄

Re: e1000e driver Network Card Detected Hardware Unit Hang

[Sirius]

In days of yore (Tue, 16 Apr 2024), Jamie thus quoth: 
> Look this is a kernel bug and Debian needs to
> fix this! Don't give me any of this crap about upstream
> this is a bug with the Debian Kernel!

Pay attention, because I am now in Support Mode as a former Principal
Technical Account Manager for Red Hat.


No, this is not necessarily a kernel bug. It can be a hardware bug and it
is plausible it can not be solved with a driver work-around.

You are hitting a problem and you want someone else to fix it for you. The
answer may simply be that you need to replace the NIC with something else.

FWIW, I have these Intel NICs in my two NUCs and they are functioning
fine. With Debian 12.5 and the latest updates.

$ lspci -v -s 00:1f.6
00:1f.6 Ethernet controller: Intel Corporation Ethernet Connection I219-V (rev 
21)
DeviceName:  LAN
Subsystem: Intel Corporation Ethernet Connection I219-V
Flags: bus master, fast devsel, latency 0, IRQ 123, IOMMU group 7
Memory at df10 (32-bit, non-prefetchable) [size=128K]
Capabilities: 
    Kernel driver in use: e1000e
Kernel modules: e1000e

The revision of the NIC may determine whether you have *hardware* problems
or not.

> This needs to be fixed!

Quick answer: replace the NIC. And do some groundwork to determine if the
NIC you replace it with has issues you should be aware of or not.

> I have already tried disabling the offloads and it does
> not work.

The specific offloads seemed to be the CRC related ones.

# ethtool -k eno1
Features for eno1:
rx-checksumming: on
tx-checksumming: on
tx-checksum-ipv4: off [fixed]
tx-checksum-ip-generic: on
tx-checksum-ipv6: off [fixed]
tx-checksum-fcoe-crc: off [fixed]
tx-checksum-sctp: off [fixed]

Note: when you disable these, throughput can drop sharply.

The other setting suggested was to hike the TX ringbuffer.

# ethtool -g eno1
Ring parameters for eno1:
Pre-set maximums:
RX: 4096
RX Mini:n/a
RX Jumbo:   n/a
TX: 4096
Current hardware settings:
RX: 256
RX Mini:n/a
RX Jumbo:   n/a
TX: 256
RX Buf Len: n/a
CQE Size:   n/a
TX Push:off
TCP data split: n/a

# ethtool -G eno1 tx 2048 rx 2048
# ethtool -g eno1
Ring parameters for eno1:
Pre-set maximums:
RX: 4096
RX Mini:n/a
RX Jumbo:   n/a
TX: 4096
Current hardware settings:
RX: 2048
RX Mini:n/a
RX Jumbo:   n/a
TX: 2048
RX Buf Len: n/a
CQE Size:   n/a
TX Push:off
TCP data split: n/a

The reason the ringbuffers are important is that the kernel and the OS can
construct packets faster in bursts than the NIC can handle, so the OS can
queue up packets in the ringbuffer and the NIC can asynchronously pick the
packets from the buffer and send them across the wire. If the ringbuffers
are set too small, they will overflow and you will get overflow errors.

# ethtool -S eno1
NIC statistics:
 rx_packets: 24463
 tx_packets: 6358
 rx_bytes: 3093199
 tx_bytes: 669733
 rx_broadcast: 8044
 tx_broadcast: 9
 rx_multicast: 10434
 tx_multicast: 2510
 rx_errors: 0
 tx_errors: 0
 tx_dropped: 0 <<<< If buffers are set too small, this increases
 multicast: 10434
 collisions: 0
 rx_length_errors: 0
 rx_over_errors: 0
 rx_crc_errors: 0
 rx_frame_errors: 0
 rx_no_buffer_count: 0
 rx_missed_errors: 0
 tx_aborted_errors: 0
 tx_carrier_errors: 0
 tx_fifo_errors: 0
 tx_heartbeat_errors: 0
 tx_window_errors: 0
 tx_abort_late_coll: 0
 tx_deferred_ok: 0
 tx_single_coll_ok: 0
 tx_multi_coll_ok: 0
 tx_timeout_count: 0
 tx_restart_queue: 0
 rx_long_length_errors: 0
 rx_short_length_errors: 0
 rx_align_errors: 0
 tx_tcp_seg_good: 0
 tx_tcp_seg_failed: 0
 rx_flow_control_xon: 9
 rx_flow_control_xoff: 9
 tx_flow_control_xon: 0
 tx_flow_control_xoff: 0
 rx_csum_offload_good: 8539<<<< If you have issues with checksum
 rx_csum_offload_errors: 0 <<<< offload, check these
 rx_header_split: 0
 alloc_rx_buff_failed: 0
 tx_smbus: 0
 rx_smbus: 0
 dropped_smbus: 0
 rx_dma_failed: 0
 tx_dma_failed: 0
 rx_hwtstamp_cleared: 0
 uncorr_ecc_errors: 0
 corr_ecc_errors: 0
 tx_hwtstamp_timeouts: 0
 tx_hwtstamp_skipped: 0

> It isn't the cable either I have tried different cables it
> still happens! This is an issue with the Kernel module for
> the e1000e NIC card.

Excellent data-point, you have ruled out whether the cable is faulty or
not. But your assumption that this is the kernel module that is broken
is still faulty.

Provably, I am running the same type of NIC (albeit a different revision)
with the same driver and I do not observe any issues. Thus, leveraging

Re: e1000e driver Network Card Detected Hardware Unit Hang

[tomas]

On Tue, Apr 16, 2024 at 09:05:29AM -0400, Stefan Monnier wrote:
> > It has been known to happen that drivers implement workarounds for issues
> > in the hardware itself, so that hardware bugs do not get tripped (or are
> > tripped less often).
> 
> 
> 
> You make it sound like it's a rare occurrence, but it's actually
> quite common.  Most of it is discrete so you'll rarely be exposed to it,
> but `grep bugs /proc/cpuinfo` is one of the places where you can see it
> being somewhat documented.

One might argue that a driver's whole raison d'être /is/ to work around
hardware bugs. But then, perhaps I'm a cynic ;-)

Cheers
-- 
t


signature.asc
Description: PGP signature

Re: e1000e driver Network Card Detected Hardware Unit Hang

[Stefan Monnier]

> It has been known to happen that drivers implement workarounds for issues
> in the hardware itself, so that hardware bugs do not get tripped (or are
> tripped less often).



You make it sound like it's a rare occurrence, but it's actually
quite common.  Most of it is discrete so you'll rarely be exposed to it,
but `grep bugs /proc/cpuinfo` is one of the places where you can see it
being somewhat documented.


Stefan

Re: e1000e driver Network Card Detected Hardware Unit Hang

[Sirius]

In days of yore (Tue, 16 Apr 2024), Sirius thus quoth: 
> In days of yore (Mon, 15 Apr 2024), Jamie thus quoth: 
> > So  there is a very nasty bug in the e1000e network card
> > driver.

Doing some reading turned up a Proxmox thread about the issues with these
Intel NICs.

https://forum.proxmox.com/threads/e1000-driver-hang.58284/page-10

May be worth scanning that thread and applying some of their solutions to
this problem.

-- 
Kind regards,

/S

Re: e1000e driver Network Card Detected Hardware Unit Hang

[Sirius]

In days of yore (Mon, 15 Apr 2024), Jamie thus quoth: 
> So  there is a very nasty bug in the e1000e network card
> driver.

https://www.intel.com/content/www/us/en/support/articles/05480/ethernet-products.html
notes that MSI interrupts may be problematic on some systems. Worth
digging into whether that is an issue on this system of yours.

I am not sure Debian can resolve this problem with the driver, but
upstream kernel folks might. Unsure whether Intel still helps maintain
this driver as it is quite old (I dealt with support issues on this driver
some 15-16 years ago).

The Intel page states this is upstream kernel only at this point, so going
to SourceForge for their out-of-tree driver is no longer an option.

> I am running Debian 12 Bookworm.
> 
> You will get the message "Detected Hardware Unit Hang" and then
> the network card just stops working.
[snip]

> This is a gigabit network card as I said it is a built in NIC I believe it
> is an Intel NIC.

It is an Intel NIC. Most of the NIC drivers beginning with an 'e' followed
by numbers are Intel as far as I know. These NICs were very common as
on-board NICs in OEM systems as Intel provided them in large volumes. They
are not the best, but they usually do their job.

[snip]
> This seems to happen when you are actually pushing a bit of traffic
> though it not a lot but just even a little bit.  It isn't network overload
> or anything I am barely doing anything really but it will do this.

If it is a hardware hang, it may be the NIC firmware getting its knickers
in a twist, and that is not something the kernel or the driver can do much
about.

> I have already tried  the following
> 
> ethtool -K eth1 tx off rx off
> ethtool -K eth1 tso off gso off
> ethtool -K eth1 gso off gro off tso off tx off rx off rxvlan off txvlan
> off sg off

All worthwhile things to try. You can also try reducing the RX buffers
from the default 4096 to 2048 if you are not running a lot of traffic. It
might not help, but worth trying.

> I have disabled all power management in the bios as well including the one
> for ASPM
> 
> I added the following to grub
> 
> pcie_aspm=off e1000e.SmartPowerDownEnable=0
> 
> 
> This is in /etc/default/grub
> GRUB_CMDLINE_LINUX_DEFAULT="quiet pcie_aspm=off
> e1000e.SmartPowerDownEnable=0"

Good thinking about power management. :)

> Then I did an update-grub as well.
> 
> None of this has worked in fixing this problem.  I am still getting the
> same issue.

Best bet at this point would be to scout the Linux Kernel Mailing List
archives to see if anyone else have run into the same problems, and then
reviewing the kernel maintainers list to find someone that works on the
e1000e driver to strike up a direct dialogue with them.

> Can you please fix this issue this is a really nasty problem with Debian
> 12 (Bookworm)
> 
> I am seeing this being reported back in Kernel 5.3.x but i am not seeing any
> reports for 6.1.x about this issue.
> 
> Debian Bug report logs - #945912
> Kernel 5.3 e100e Detected Hardware Unit Hang
> https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=945912

If it has been reported before and is still present now, one of two things
is likely true.
 1) the problem was intermittent and could not be reliably reproduced in
order to debug and resolve
 2) the problem was related to the hardware itself, and there was no way
to fix it in either driver or firmware

It has been known to happen that drivers implement workarounds for issues
in the hardware itself, so that hardware bugs do not get tripped (or are
tripped less often).

> Please reply back and confirm that you got this email and that you are
> looking into this problem please.

To state the obvious, I am not a kernel maintainer for Debian and do not
speak on behalf of the Debian project.

I work for a Linux company you may have heard of and have done so for
almost eighteen years, a decade of which was in support. 15 years ago, I
know exactly who I would have gone to to look into this problem, but he
now works for Broadcom and probably has forgotten all about the
e1000/e1000e drivers.

Upstream driver maintainer would be the best bet IMHO. If this driver is
community support only (i.e. if Intel no longer participates in driver
maintenance), I would say that all bets are off.

With only one datapoint - your system and your NIC, it is not possible to
rule out that the NIC itself is bad. :-/

> -- This email message, including any attachments, is for the intended
> recipient(s) only and may contain information that is privileged,
> confidential and/or exempt from disclosure under applicable law. If you
> have received this message in error, or are obviously not one of the
> intended recipients, please immediately notify the sender by reply email
> and delete this email message, including any attach

Re: Help to report a bug related to a usb3 lan adapter driver

[Charles Curley]

On Mon, 15 Apr 2024 20:57:00 +0200
user7415 same  wrote:

> I had a discussion in stack exchange related to the problem that is
> well explained here:
> https://unix.stackexchange.com/questions/774594/debian-12-all-of-sudden-my-usb3-lan-adapter-get-assigned-random-mac-address-ea
> 
> For what I understood the problem was fixed in 6.8, but I'm using
> debian 12 that will never use that so much new kernel I guess, could
> you help me to report officially the bug so that the upstream channel
> will correct it by the 6.1.0-22 version ?

Bookwom backports has linux-image-6.6.13+bpo-amd64. You might try that.
https://backports.debian.org/

It just so happens I have one of the same beasties. I just plugged it
in to a machine running kernel 6.6.13+bpo-amd64, unplugged it, waited
20 seconds, and plugged it in to another machine running kernel
6.5.0-0.deb12.4-amd64. I then plugged it into a machine with
6.6.13+bpo-amd64. All three times I got a MAC address of
8c:ae:4c:d6:22:17. So either of those kernels might well work for you.

-- 
Does anybody read signatures any more?

https://charlescurley.com
https://charlescurley.com/blog/

e1000e driver Network Card Detected Hardware Unit Hang

[Jamie]

So  there is a very nasty bug in the e1000e network card
driver.

I am running Debian 12 Bookworm.

You will get the message "Detected Hardware Unit Hang" and then
the network card just stops working.

This is a built in NIC  on the computer
The computer is a is a HP Prodesk 600 G4 MT

This is the mini tower version as denoted by the MT.


This log comes from my /var/log/syslog.


Apr 15 01:57:12 gateway vmunix: [ 7743.893557] e1000e :00:1f.6 eth1: 
Detected Hardware Unit Hang:

Apr 15 01:57:12 gateway vmunix: [ 7743.893557] TDH  
Apr 15 01:57:12 gateway vmunix: [ 7743.893557] TDT  
Apr 15 01:57:12 gateway vmunix: [ 7743.893557] next_to_use  
Apr 15 01:57:12 gateway vmunix: [ 7743.893557] next_to_clean    
Apr 15 01:57:12 gateway vmunix: [ 7743.893557] buffer_info[next_to_clean]:
Apr 15 01:57:12 gateway vmunix: [ 7743.893557] time_stamp   
<1001c6345>

Apr 15 01:57:12 gateway vmunix: [ 7743.893557] next_to_watch    
Apr 15 01:57:12 gateway vmunix: [ 7743.893557] jiffies  
<1001c6550>

Apr 15 01:57:12 gateway vmunix: [ 7743.893557] next_to_watch.status <0>
Apr 15 01:57:12 gateway vmunix: [ 7743.893557] MAC Status 
<80083>

Apr 15 01:57:12 gateway vmunix: [ 7743.893557] PHY Status <796d>
Apr 15 01:57:12 gateway vmunix: [ 7743.893557] PHY 1000BASE-T Status  <3800>
Apr 15 01:57:12 gateway vmunix: [ 7743.893557] PHY Extended Status    <3000>
Apr 15 01:57:12 gateway vmunix: [ 7743.893557] PCI Status <10>
Apr 15 01:57:13 gateway vmunix: [ 7744.123237] net-fw DROP IN=eth0 OUT= 
MAC=00:13:3b:e3:8f:b0:0c:a4:02:35:6d:87:08:00 SRC=75.159.223.219 
DST=199.126.41.116 LE>
Apr 15 01:57:13 gateway vmunix: [ 7744.417235] net-fw DROP IN=eth0 OUT= 
MAC=00:13:3b:e3:8f:b0:0c:a4:02:35:6d:87:08:00 SRC=75.159.223.219 
DST=199.126.41.116 LE>
Apr 15 01:57:14 gateway vmunix: [ 7745.412183] net-fw DROP IN=eth0 OUT= 
MAC=00:13:3b:e3:8f:b0:0c:a4:02:35:6d:87:08:00 SRC=75.159.223.219 
DST=199.126.41.116 LE>
Apr 15 01:57:14 gateway vmunix: [ 7745.659234] net-fw DROP IN=eth0 OUT= 
MAC=00:13:3b:e3:8f:b0:0c:a4:02:35:6d:87:08:00 SRC=75.159.223.219 
DST=199.126.41.116 LE>
Apr 15 01:57:14 gateway vmunix: [ 7745.877564] e1000e :00:1f.6 eth1: 
Detected Hardware Unit Hang:

Apr 15 01:57:14 gateway vmunix: [ 7745.877564] TDH  
Apr 15 01:57:14 gateway vmunix: [ 7745.877564] TDT  
Apr 15 01:57:14 gateway vmunix: [ 7745.877564] next_to_use  
Apr 15 01:57:14 gateway vmunix: [ 7745.877564] next_to_clean    
Apr 15 01:57:14 gateway vmunix: [ 7745.877564] buffer_info[next_to_clean]:
Apr 15 01:57:14 gateway vmunix: [ 7745.877564] time_stamp   
<1001c6345>

Apr 15 01:57:14 gateway vmunix: [ 7745.877564] next_to_watch    
Apr 15 01:57:14 gateway vmunix: [ 7745.877564] jiffies  
<1001c6740>

Apr 15 01:57:14 gateway vmunix: [ 7745.877564] next_to_watch.status <0>
Apr 15 01:57:14 gateway vmunix: [ 7745.877564] MAC Status 
<80083>

Apr 15 01:57:14 gateway vmunix: [ 7745.877564] PHY Status <796d>
Apr 15 01:57:14 gateway vmunix: [ 7745.877564] PHY 1000BASE-T Status  <3800>
Apr 15 01:57:14 gateway vmunix: [ 7745.877564] PHY Extended Status    <3000>
Apr 15 01:57:14 gateway vmunix: [ 7745.877564] PCI Status <10>
Apr 15 01:57:15 gateway vmunix: [ 7746.220253] net-fw DROP IN=eth0 OUT= 
MAC=00:13:3b:e3:8f:b0:0c:a4:02:35:6d:87:08:00 SRC=75.159.223.219 
DST=199.126.41.116 LE>
Apr 15 01:57:15 gateway vmunix: [ 7746.485268] net-fw DROP IN=eth0 OUT= 
MAC=00:13:3b:e3:8f:b0:0c:a4:02:35:6d:87:08:00 SRC=75.159.223.219 
DST=199.126.41.116 LE>
Apr 15 01:57:16 gateway vmunix: [ 7747.893578] e1000e :00:1f.6 eth1: 
Detected Hardware Unit Hang:

Apr 15 01:57:16 gateway vmunix: [ 7747.893578] TDH  
Apr 15 01:57:16 gateway vmunix: [ 7747.893578] TDT  
Apr 15 01:57:16 gateway vmunix: [ 7747.893578] next_to_use  
Apr 15 01:57:16 gateway vmunix: [ 7747.893578] next_to_clean    
Apr 15 01:57:16 gateway vmunix: [ 7747.893578] buffer_info[next_to_clean]:
Apr 15 01:57:16 gateway vmunix: [ 7747.893578] time_stamp   
<1001c6345>

Apr 15 01:57:16 gateway vmunix: [ 7747.893578] next_to_watch    
Apr 15 01:57:16 gateway vmunix: [ 7747.893578] jiffies  
<1001c6938>

Apr 15 01:57:16 gateway vmunix: [ 7747.893578] next_to_watch.status <0>
Apr 15 01:57:16 gateway vmunix: [ 7747.893578] MAC Status 
<80083>

Apr 15 01:57:16 gateway vmunix: [ 7747.893578] PHY Status <796d>
Apr 15 01:57:16 gateway vmunix: [ 7747.893578] PHY 1000BASE-T Status  <3800>
Apr 15 01:57:16 gateway vmunix: [ 7747.893578] PHY Extended Status    <3000>
Apr 15 01:57:16 gateway vmunix: [ 7747.893578] PCI Status <10>


It does this multiple tim

Help to report a bug related to a usb3 lan adapter driver

[user7415 same]

Hello,

I'm not very skilled in what concerns to kernel patching and compiling.

I had a discussion in stack exchange related to the problem that is well
explained here:
https://unix.stackexchange.com/questions/774594/debian-12-all-of-sudden-my-usb3-lan-adapter-get-assigned-random-mac-address-ea

For what I understood the problem was fixed in 6.8, but I'm using debian 12
that will never use that so much new kernel I guess, could you help me to
report officially the bug so that the upstream channel will correct it by
the 6.1.0-22 version ?

Thank you very much!

Re: Where to report print driver bug

[Marco Moock]

Am Fri, 23 Feb 2024 14:47:41 -0500
schrieb James Klaas :

> "Generic PCL 6/PCL XL Printer Foomatic/pxlcolor (recommended)"

Do you know the file that provides that?
If so, you apt-file search "file" to find the package that provides it.

Where to report print driver bug

[James Klaas]

I was going to submit a bug for this but I don't know what package I 
should report the bug against.


Debian bugreport says:
Please enter the name of the package in which you have found a problem, 
or type 'other' to report a more general problem. If you don't know what 
package the bug is in, please contact debian-user@lists.debian.org for 
assistance.


I have a Dell 2130cn, which is a PCL6 compatible printer. CUPS/Print 
manager says


"Generic PCL 6/PCL XL Printer Foomatic/pxlcolor (recommended)"

is the recommended driver and the only PCL6 driver that actually prints 
in color as far as I can tell. However, sometime in the past year or two 
(maybe between Bullseye and Bookworm), this driver no longer allows me 
to print in duplex.


I was able to print in duplex from Windows, so I decided to try a 
different driver. I tried both


"Generic PCL 6/PCL XL Printer - CUPS+Gutenprint v5.3.4"

and

"Generic PCL 6/PCL XL Printer Foomatic/hpijs-pcl5c"

They both print in duplex, but do not print in color.

I would like to submit a bug report, but I do not know if I should only 
submit one for printer-driver-pxljr, which I think provides pxlcolor or 
something more generic.

Re: Nvidia driver on Debian 10

[Franco Martelli]

On 03/01/24 at 19:28, Thomas Anderson wrote:

Thanks a lot, Macro!

it worked. Sorry for delay, I got side tracked with "my day" =)

On 03/01/2024 12:03, Marco Moock wrote:

Am 03.01.2024 um 11:52:59 Uhr schrieb Thomas Anderson:


Here is the output. Wow, a lot.

All those lines with ii at the beginning mean the package is installed.

apt remove *nvidia* -s

Check if the result is ok and then run it without -s (-s only
simulates).





Give it a try to nouveau driver, here it works very well with kernel 
6.1.55 the only drawback I have is to restart the windows compositor 
when I suspend to RAM, good luck


Cheers
--
Franco Martelli

Re: Nvidia driver on Debian 10

[jeremy ardley]

On 3/1/24 18:36, Thomas Anderson wrote:
I will upgrade Debian to 12 within the next month or so...but before I 
do, want to clean up my current system.



I am running Debian 12 with the nvidia driver. It's mostly OK but I 
recently had to reboot as the Xorg process had taken around 70% of my 
32G Ram. This may be a driver problem?


I have other issues with applications including firefox-esr and chrome 
taking lots of memory - though this is unlikely to be related to the 
video driver.

Re: Nvidia driver on Debian 10

[Thomas Anderson]

Thanks a lot, Macro!

it worked. Sorry for delay, I got side tracked with "my day" =)

On 03/01/2024 12:03, Marco Moock wrote:

Am 03.01.2024 um 11:52:59 Uhr schrieb Thomas Anderson:


Here is the output. Wow, a lot.

All those lines with ii at the beginning mean the package is installed.

apt remove *nvidia* -s

Check if the result is ok and then run it without -s (-s only
simulates).

Re: Nvidia driver on Debian 10

[Hans]

> dpkg-reconfigure nvidia-kernel-dkms
> 

Hi all,

I am using a NVidia card, which is using the Tesla-driver version (525.147.XX) 
and used the 
470.XXX version before.

Both could not be installed by using the above package.

But it got installed using package module-assistant.

Suggestion, how to fix:

First of, move any /usr/bin/gdm|kdm|xdm somewhere else, so the graphical login 
manager will 
not start.

Secondly, install package "module-assistant".

Third: Run it, command is "m-a", its starts an ncurses surface. 

Fourth: go through all the processes: "update" , then "prepare" (which will 
download all needed 
apps and files like compiler, kernel-headers and so on), then go to "select". 

Chose your needed driver version.  

Then again, chose "get", which is downloading the NVidia-packages.

Then chose "build" and it will automatiucally create a debian package, which 
then will be 
installed with the option "install".

After it exit it.

Now: First aof all reboot!

If rebooted, check now with the command "startx" as root, if X is starting. If 
yes, mode /usr/bin/
xdm|kdm|sddm|gdm back.

Reboot now again, Voila!



Hints: If it is NOT starting, check if the nouveau driver is still loaded. If 
yes, try to reboot, if it will 
be loaded again, try to blacklist it.

If blacklisting is NOT workin, do the hard way , an delete or remove the 
nouveau.ko from the 
kernel tree.



If X is still NOT starting, check xorg.logs to look for reasons.



Second hint: IIn the past, I got problems with wrong versions. Nvidia said, I 
should use i.e. 
390xx, but in real, I had to use 340xx. So try also one version prior.




Hope this helps!

Godd luck and

best regards

Hans

Re: Nvidia driver on Debian 10

[Anssi Saari]

Marco Moock  writes:

> Am 03.01.2024 um 11:52:59 Uhr schrieb Thomas Anderson:
>
>> Here is the output. Wow, a lot.
>
> All those lines with ii at the beginning mean the package is installed.
>
> apt remove *nvidia* -s
>
> Check if the result is ok and then run it without -s (-s only
> simulates).

nvidia-driver is the actual top level metapackage so removing that is
likely the thing. Although I'm not sure that makes much sense if Thomas
gets just a blank screen when booting the kernel without the nvidia
drivers.

Also I think fixing the drivers for both installed kernels might be as
simple as running

dpkg-reconfigure nvidia-kernel-dkms

The version of nvidia-driver and related packages seem to be the same as
what nvidia-smi reports, 418.226.00, so it doesn't look like there's
anything super weird in the system.

Re: Nvidia driver on Debian 10

[Marco Moock]

Am 03.01.2024 um 11:52:59 Uhr schrieb Thomas Anderson:

> Here is the output. Wow, a lot.

All those lines with ii at the beginning mean the package is installed.

apt remove *nvidia* -s

Check if the result is ok and then run it without -s (-s only
simulates).

Re: Nvidia driver on Debian 10

[Thomas Anderson]

Here is the output. Wow, a lot.

Desired=Unknown/Install/Remove/Purge/Hold
| 
Status=Not/Inst/Conf-files/Unpacked/halF-conf/Half-inst/trig-aWait/Trig-pend 


|/ Err?=(none)/Reinst-required (Status,Err: uppercase=bad)
||/ Name Version Architecture Description
+++-==--- 


un bumblebee-nvidia (no description available)
ii glx-alternative-nvidia 1.0.0 amd64 allows the selection of NVIDIA as 
GLX provider

un libegl-nvidia-legacy-390xx0 (no description available)
un libegl-nvidia-tesla-418-0 (no description available)
ii libegl-nvidia0:amd64 418.226.00-3 amd64 NVIDIA binary EGL library
ii libegl-nvidia0:i386 418.226.00-3 i386 NVIDIA binary EGL library
un libegl1-glvnd-nvidia (no description available)
un libegl1-nvidia (no description available)
un libgl1-glvnd-nvidia-glx (no description available)
ii libgl1-nvidia-glvnd-glx:amd64 418.226.00-3 amd64 NVIDIA binary 
OpenGL/GLX library (GLVND variant)
ii libgl1-nvidia-glvnd-glx:i386 418.226.00-3 i386 NVIDIA binary 
OpenGL/GLX library (GLVND variant)

un libgl1-nvidia-glx (no description available)
un libgl1-nvidia-glx-418.226.00 (no description available)
un libgl1-nvidia-glx-any (no description available)
un libgl1-nvidia-glx-i386 (no description available)
un libgl1-nvidia-legacy-390xx-glx (no description available)
un libgl1-nvidia-tesla-418-glx (no description available)
un libgldispatch0-nvidia (no description available)
ii libgles-nvidia1:amd64 418.226.00-3 amd64 NVIDIA binary OpenGL|ES 1.x 
library
ii libgles-nvidia1:i386 418.226.00-3 i386 NVIDIA binary OpenGL|ES 1.x 
library
ii libgles-nvidia2:amd64 418.226.00-3 amd64 NVIDIA binary OpenGL|ES 2.x 
library
ii libgles-nvidia2:i386 418.226.00-3 i386 NVIDIA binary OpenGL|ES 2.x 
library

un libgles1-glvnd-nvidia (no description available)
un libgles2-glvnd-nvidia (no description available)
un libglvnd0-nvidia (no description available)
ii libglx-nvidia0:amd64 418.226.00-3 amd64 NVIDIA binary GLX library
ii libglx-nvidia0:i386 418.226.00-3 i386 NVIDIA binary GLX library
un libglx0-glvnd-nvidia (no description available)
ii libnvidia-cbl:amd64 418.226.00-3 amd64 NVIDIA binary Vulkan ray 
tracing (cbl) library

un libnvidia-cbl-418.226.00 (no description available)
un libnvidia-cfg.so.1 (no description available)
ii libnvidia-cfg1:amd64 418.226.00-3 amd64 NVIDIA binary OpenGL/GLX 
configuration library

un libnvidia-cfg1-any (no description available)
ii libnvidia-eglcore:amd64 418.226.00-3 amd64 NVIDIA binary EGL core 
libraries
ii libnvidia-eglcore:i386 418.226.00-3 i386 NVIDIA binary EGL core 
libraries

un libnvidia-eglcore-418.226.00 (no description available)
ii libnvidia-encode1:amd64 418.226.00-3 amd64 NVENC Video Encoding 
runtime library
ii libnvidia-encode1:i386 418.226.00-3 i386 NVENC Video Encoding runtime 
library
ii libnvidia-fatbinaryloader:amd64 418.226.00-3 amd64 NVIDIA FAT binary 
loader
ii libnvidia-fatbinaryloader:i386 418.226.00-3 i386 NVIDIA FAT binary 
loader

un libnvidia-fatbinaryloader-418.226.00 (no description available)
ii libnvidia-glcore:amd64 418.226.00-3 amd64 NVIDIA binary OpenGL/GLX 
core libraries
ii libnvidia-glcore:i386 418.226.00-3 i386 NVIDIA binary OpenGL/GLX core 
libraries

un libnvidia-glcore-418.226.00 (no description available)
ii libnvidia-glvkspirv:amd64 418.226.00-3 amd64 NVIDIA binary Vulkan 
Spir-V compiler library
ii libnvidia-glvkspirv:i386 418.226.00-3 i386 NVIDIA binary Vulkan 
Spir-V compiler library

un libnvidia-glvkspirv-418.226.00 (no description available)
un libnvidia-legacy-340xx-cfg1 (no description available)
un libnvidia-legacy-390xx-cfg1 (no description available)
un libnvidia-ml.so.1 (no description available)
ii libnvidia-ml1:amd64 418.226.00-3 amd64 NVIDIA Management Library 
(NVML) runtime library
ii libnvidia-ptxjitcompiler1:amd64 418.226.00-3 amd64 NVIDIA PTX JIT 
Compiler library
ii libnvidia-ptxjitcompiler1:i386 418.226.00-3 i386 NVIDIA PTX JIT 
Compiler library
ii libnvidia-rtcore:amd64 418.226.00-3 amd64 NVIDIA binary Vulkan ray 
tracing (rtcore) library

un libnvidia-rtcore-418.226.00 (no description available)
un libopengl0-glvnd-nvidia (no description available)
ii nvidia-alternative 418.226.00-3 amd64 allows the selection of NVIDIA 
as GLX provider

un nvidia-alternative--kmod-alias (no description available)
un nvidia-alternative-legacy-173xx (no description available)
un nvidia-alternative-legacy-71xx (no description available)
un nvidia-alternative-legacy-96xx (no description available)
un nvidia-cuda-mps (no description available)
un nvidia-current (no description available)
un nvidia-current-updates (no description available)
ii nvidia-detect 418.226.00-3 amd64 NVIDIA GPU detection utility
ii nvidia-driver 418.226.00-3 amd64 NVIDIA metapackage
un nvidia-driver-any (no description available)
ii nvidia-driver-bin 418.226.00-3 amd64 NVIDIA driver support binaries
un nvidia-driver-bin-418.226.00

Re: Nvidia driver on Debian 10

[Marco Moock]

Am 03.01.2024 um 11:36:03 Uhr schrieb Thomas Anderson:

> Key information:
> I am aware that there are two different installation paths for the 
> nvidia driver, either through the repos or from nvidia downloads.
> I think I did the later, but I have misplaced, lost, and cannot
> download the driver listed above to uninstall it. Apparently, you
> need the nvidia installer to uninstall the driver.

> I tried from CLI:
> apt nvidia purge
> but that did not work.

Because that only applies to DEB packages installed (most of them come
from a repository).
If you installed it by the shell script, it is unrelated to the
apt/dpkg stuff.

Run dpkg -l *nvidia*

and show the output.

Nvidia driver on Debian 10

[Thomas Anderson]

Ok, throwing up the bat signal.

I will upgrade Debian to 12 within the next month or so...but before I 
do, want to clean up my current system.


I have to kernels:

4.19.0-25 and 4.19.0-24.

The latter will boot fine, the former (newer kernel) will not. it just 
goes to a blinking cursor, indefinitely (I guess, didn't wait for infinity).


I double checked the space on the boot partition, but it's fine, with 
25% space still left.


when i type:

nvidia-smi

from the CLI, I get...

+-+
|NVIDIA-SMI 418.226.00   Driver Version: 418.226.00   CUDA Version: 10.1 
    |

|---+--+--+
|GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. 
ECC |
|Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util 
 Compute M. |

|===+==+==|
|  0  GeForce GT 1030     On   | :1F:00.0  On |                 
 N/A |
|N/A   46C    P8    N/A /  19W |    340MiB /  2000MiB |     39%     
 Default |

+---+--+--+
+-+
|Processes:                                                       GPU 
Memory |
| GPU       PID   Type   Process name                             Usage 
     |

|=|
|   0      2107      G   /usr/lib/xorg/Xorg                           
259MiB |
|   0      4220      G   cinnamon                                     
 57MiB |
|   0     11718      G   nvidia-settings                               
 0MiB |
|   0     11745      G   nvidia-settings                               
 0MiB |

Key information:
I am aware that there are two different installation paths for the 
nvidia driver, either through the repos or from nvidia downloads.
I think I did the later, but I have misplaced, lost, and cannot download 
the driver listed above to uninstall it. Apparently, you need the nvidia

installer to uninstall the driver.
I tried from CLI:
apt nvidia purge
but that did not work.
As a last gasp of desperation, I tried to ignore the warning from the 
new current nvidia downloaded installer, and just install anyway, but it 
also

does not install the driver.
Reading logs, it says I have another version driver. When I ran that 
file with the -uninstall flag, it instead said I did not have

an nvidia driver installed.
In summation:
1. I apparently have an nvidia driver installed (proof with nvidia-smi)
2. How it was installed - I am guessing from download file not repos, 
but not 100% sure.

3. I know linux and nvidia go together like oil and water...
Since not really using this for gaming or anything more than using two 
monitors, I might just shelf it and delete the latest kernel and set
the older kernel as the default booting kernel--for now, and setup my 
server properly in the future.
Anyone offer up a silver bullet solution that would be nice =) 
Otherwise, just gonna shelf it for now.

Re: Re: RTL8852 driver for Debian 11

[Timothy M Butterworth]

On Sun, Nov 26, 2023 at 2:11 AM Andy Smith  wrote:

> Hi,
>
> Again ignoring your more non-constructive complaints…
>
> TL;DR: Try USB networking like by plugging in your phone or a USB
> ethernet/wifi dongle.
>
> On Sat, Nov 25, 2023 at 07:39:20PM +, Richard Smith wrote:
> > The only place I found that had the driver required me to install
> > more content from online (which I can't access, due to the laptop
> > not having an Ethernet port, as well as no wireless).
>
> It is not unusual for computers (especially laptops) with newer
> hardware to not have needed drivers contained in the distro kernel.
> It's worth thinking about how you will handle that before you buy
> it.
>
> One valid strategy for handling that is, "I will consider just using
> Ubuntu, or Fedora, or whatever, and only buy hardware that works in
> those distributions."
>
> Assuming you want to continue with a distribution that doesn't
> ship a kernel that supports your wifi, and you need to get files
> onto the computer, you will have to get a bit more creative.
>
> As mentioned in the other thread, I have a laptop that has such a
> wifi card. I solved the problem of no initial networking by plugging
> a USB cable from my phone to the laptop. NetworkManager then offered
> to use it as a USB network connection without me having to configure
> anything at all. It was a one click temporary solution to getting
> the wifi driver DKMS and everything needed to compile it.
>
> Other possibilities off the top of my head:
>
> - USB ethernet dongle
>
> - USB wifi dongle with a supported chipset
>

I got burned on my last Laptop Upgrade. No WiFi, basic GPU etc. I upgraded
to Debian testing to get the latest drivers and all was good. I upgraded
straight through testing to Bookworm stable-new. It sounds like you may
want to do the same thing with Trixie. I bought one of these, a little
pricey but having it on hand to get my system up and running was worth
every penny. It is not the latest and greatest but it runs on all free
firmware!

https://www.thinkpenguin.com/gnu-linux/penguin-wireless-n-usb-adapter-gnu-linux-tpe-n150usb




> - File transfer by USB
>
> I'm sure you and others can think of more.
>
> > I am spending WAY too much time installing this - with very little
> > to show for it!
>
> Don't use it then. Non-actionable complaints are difficult to
> address in a volunteer project and it's totally fine for you to use
> something else. There are certainly things I do not use Debian (or
> Linux in general) for that while they would be technically possible,
> are just "way too much time" for me to consider worth it. The answer
> isn't for me to drop my complaints at Linux's door.
>
> Thanks,
> Andy
>
> --
> https://bitfolk.com/ -- No-nonsense VPS hosting
>
>

-- 
⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org/
⠈⠳⣄⠀⠀

Re: Re: RTL8852 driver for Debian 11

[Andy Smith]

Hi,

Again ignoring your more non-constructive complaints…

TL;DR: Try USB networking like by plugging in your phone or a USB
ethernet/wifi dongle.

On Sat, Nov 25, 2023 at 07:39:20PM +, Richard Smith wrote:
> The only place I found that had the driver required me to install
> more content from online (which I can't access, due to the laptop
> not having an Ethernet port, as well as no wireless).

It is not unusual for computers (especially laptops) with newer
hardware to not have needed drivers contained in the distro kernel.
It's worth thinking about how you will handle that before you buy
it.

One valid strategy for handling that is, "I will consider just using
Ubuntu, or Fedora, or whatever, and only buy hardware that works in
those distributions."

Assuming you want to continue with a distribution that doesn't
ship a kernel that supports your wifi, and you need to get files
onto the computer, you will have to get a bit more creative.

As mentioned in the other thread, I have a laptop that has such a
wifi card. I solved the problem of no initial networking by plugging
a USB cable from my phone to the laptop. NetworkManager then offered
to use it as a USB network connection without me having to configure
anything at all. It was a one click temporary solution to getting
the wifi driver DKMS and everything needed to compile it.

Other possibilities off the top of my head:

- USB ethernet dongle

- USB wifi dongle with a supported chipset

- File transfer by USB

I'm sure you and others can think of more.

> I am spending WAY too much time installing this - with very little
> to show for it!

Don't use it then. Non-actionable complaints are difficult to
address in a volunteer project and it's totally fine for you to use
something else. There are certainly things I do not use Debian (or
Linux in general) for that while they would be technically possible,
are just "way too much time" for me to consider worth it. The answer
isn't for me to drop my complaints at Linux's door.

Thanks,
Andy

-- 
https://bitfolk.com/ -- No-nonsense VPS hosting

Re: RTL8852 driver for Debian 11

[Andy Smith]

Hi,

On Sat, Nov 25, 2023 at 12:21:28PM +, Richard Smith wrote:
> I find the issue with Debian weird, as I was able to use this same computer 
> (Lenovo IdeaPad I7, with a Realtek b52 controller) with Ubuntu, Kali, and 
> Tails, without an issue!
> 
> I hear that Debian is awesome when it comes to stability - but after spending 
> over 8 hours trying to install it to my laptop, I'm beginning to question 
> what constitutes as "awesome" in some people's eyes...

Your opinion piece is best posted to your blog as it won't help you
get assistance in a Debian support venue.

> I'm trying to learn about how to compile a driver (if I can find
> one for Debian 12) - if anyone can assist, that would be great!

I have a laptop with an RTL8852be which runs Debian 12. I had to
install a DKMS module to get support for it. I reported this to
debian-kernel here:

https://lists.debian.org/debian-kernel/2023/03/msg00275.html

that link also contains a link to the github repository for the DKMS
that I successfully use. If that driver is for your hardware, I
suggest you try that and report back any issues. It works fine for
me and my only complaint is that it is a DKMS.

Or continue using a distro kernel that has the driver built in.

As the thread linked above covers, given that upstream kernel does
have the rtw89_8852be driver in it, it will eventually come to
Debian, though possibly in a bookworm-backports kernel untilDebian
13 (trixie) is released.

Thanks,
Andy

-- 
https://bitfolk.com/ -- No-nonsense VPS hosting

Re: Re: RTL8852 driver for Debian 11

[Richard Smith]

I'm currently wrestling with Debian 12 for my Lenovo IdeaPad 7 laptop - after 
having the NIC function swimmingly with Ubuntu, Kali, LMDE, and Tails.
Not only is wireless not seen - but it seems that the functionality of the DVD 
version leaves a LOT to be desired. The only place I found that had the driver 
required me to install more content from online (which I can't access, due to 
the laptop not having an Ethernet port, as well as no wireless).

When one couples the above with the Debian site being horrible organized 
(seriously: it gives me the impression that people don't want you to use 
Debian), I cant help but think: for all of this talk about Debian being 
awesome, I am spending WAY too much time installing this - with very little to 
show for it!

signature.asc
Description: OpenPGP digital signature

Re: RTL8852 driver for Debian 11

[David Wright]

On Sat 25 Nov 2023 at 12:21:28 (+), Richard Smith wrote:
> Testing access to the message.
> 
> I find the issue with Debian weird, as I was able to use this same computer 
> (Lenovo IdeaPad I7, with a Realtek b52 controller) with Ubuntu, Kali, and 
> Tails, without an issue!

In ubuntu, cat /proc/modules would tell you the modules loaded,
and   dmesg | grep firmware   would tell you the firmware required.

> I hear that Debian is awesome when it comes to stability - but after spending 
> over 8 hours trying to install it to my laptop, I'm beginning to question 
> what constitutes as "awesome" in some people's eyes...
> 
> I'm trying to learn about how to compile a driver (if I can find one for 
> Debian 12) - if anyone can assist, that would be great!

Is this an X-Y problem, and you don't really want to compile one,
but just find it.

Cheers,
David.

Re: RTL8852 driver for Debian 11

[Charles Curley]

On Sat, 25 Nov 2023 12:21:28 +
Richard Smith  wrote:

> I find the issue with Debian weird, as I was able to use this same
> computer (Lenovo IdeaPad I7, with a Realtek b52 controller) with
> Ubuntu, Kali, and Tails, without an issue!

What exactly is your problem? Please show error messages, and also run
lspci or lsusb as appropriate and show us the result.

I searched thinkwiki (https://www.thinkwiki.org/wiki/ThinkWiki) for
"IdeaPad I7" and got nothing.

"RTL8852" is not sufficient to identify which firmware package you need.

root@tsalmoth:~# apt show firmware-realtek | grep -i  RTL8852

WARNING: apt does not have a stable CLI interface. Use with caution in scripts.

  * Realtek RTL8852AU Bluetooth config (rtl_bt/rtl8852au_config.bin)
  * Realtek RTL8852AU Bluetooth firmware (rtl_bt/rtl8852au_fw.bin)
  * Realtek RTL8852BU Bluetooth config (rtl_bt/rtl8852bu_config.bin)
  * Realtek RTL8852BU Bluetooth firmware (rtl_bt/rtl8852bu_fw.bin)
  * Realtek RTL8852CU Bluetooth config (rtl_bt/rtl8852cu_config.bin)
  * Realtek RTL8852CU Bluetooth firmware (rtl_bt/rtl8852cu_fw.bin)
  * Realtek RTL8852A firmware, version v0.9.12.2 (rtw89/rtw8852a_fw.bin)
  * Realtek RTL8852B firmware, version v0.27.32.1
  * Realtek RTL8852C firmware, version v0.27.56.10
root@tsalmoth:~# 

Running lspci should give you the exact designation of the device you
have, possibly something similar to this:

root@hawk:~# lspci | grep -i realtek
03:00.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL8111/8168/8411 
PCI Express Gigabit Ethernet Controller (rev 11)
root@hawk:~# 

Please run an appropriate command, then copy and paste into your reply
email, including leading and trailing shell prompts.

You might run the same command(s) under one of the Linux distributions
that does work, and show us those results as well.


-- 
Does anybody read signatures any more?

https://charlescurley.com
https://charlescurley.com/blog/

Re: RTL8852 driver for Debian 11

[Richard Smith]

Testing access to the message.

I find the issue with Debian weird, as I was able to use this same computer 
(Lenovo IdeaPad I7, with a Realtek b52 controller) with Ubuntu, Kali, and 
Tails, without an issue!

I hear that Debian is awesome when it comes to stability - but after spending 
over 8 hours trying to install it to my laptop, I'm beginning to question what 
constitutes as "awesome" in some people's eyes...

I'm trying to learn about how to compile a driver (if I can find one for Debian 
12) - if anyone can assist, that would be great!



signature.asc
Description: OpenPGP digital signature

More printers in the Brother printer-driver-brlaser package

[C.T.F. Jansen]

Greetings,

There are drivers for more Brother laser printers in the package 
printer-driver-brlaser .


One doesn't have to put up with Brother's atrocious install script to 
set up the DCP-1610W.
The driver package from the Debian repository doesn't produce the 
extraneous squiggles and blobs either. Not so far.


Good luck.

frank.jan...@actrix.gen.nz ,  ZL2TTS

Re: Amanda (was: sata driver compataility Q)

[Glenn]

My all time fav for file level backups is BackupPC. De-duplicates files in the 
compressed pool.

Glenn 

On September 19, 2023 12:18:36 p.m. ADT, gene heskett  
wrote:
>On 9/19/23 08:59, Stefan Monnier wrote:
>>> Compared to the setup required for amanda, that sounds very inviting. Amanda
>>> has a very steep learning curve just because it is so versatile I'm still
>>> waiting on stuff, so no more actual progress.
>> 
>> I used Amanda many years ago and was quite pleased with it, but I must
>> say I'm having a hard time imagining it in my current world where tapes
>> don't make much sense for backups.
>> 
>Thats where vtapes come in, A vtape is nothing more than a directory on the 
>backup medium, which for me was a BIG hard drive with in my case, 60 suddirs, 
>used as tapes. Each contained individual files identified as to backup level 
>which was a way ti differentiate a full copy, or what had ben changed since 
>the last full, or what had been changed since the last level 1, wash rinse 
>repeat for ever deeper levels. And with or w/o compression. Executables 
>generally aren't worth the time to compress. Ditto for a dir full if pictures 
>or pdf's. They are not very compressible.  In the days of tapes, a buffer 
>drive was used to build up each entry as a big file that was then copied to 
>the tape w/o any shoe-shining of the tapedrive, saveibg the huge wear and tear 
>of the tape if it had to stop and wait for data from the compressor. Then back 
>uo a few feet, back forward to begin a fresh write at the end of the previous 
>track. But since spinning rust is random access, amd so is the vtape, I don't 
>think the anti-shoeshine has much if any advantage whrn using vtapes. With 
>some filesystems it might reduce fragmentation but that was never a problem 
>with ext4. I ran with that buffer drive for about 17 yeas, starting out with a 
>4 tape seagate dds4 tape drive but it was by far, the least dependable thing 
>in that whole chain. I was then backing up 3 cnc machines and this ones 
>predecessor, but the drive needed a months vacation in Oklahoma city about 2x 
>a year for a new head drum that seagate would not sell me, a CET with 
>extensive experience replacing even smaller, more precise and damn sure more 
>expensive at $3500 a copy dvc-pro broadcast vcr heads.  So I tried vtapes, 
>first on a 220G drive but soon opted for a bigger one as they became 
>available, and had just graduated to a pair of seagates first 2T's, both of 
>which just disapeared off the sata buss in the middle of the night. the main 
>drive for this machine and the amanda drive. They were about 2 weeks old. So I 
>rebuilt this machine using a 500G Samsung SSD. I was out of the amanda 
>business and lost everything with those 2 failures, whih upset me so much I 
>never tried to warranty them. I was done with spinning rust.
>
>Some of the loss was the only pix of my first wife who had a stroke and died 
>in '68 at 34. Left me with 3 children to raise, but the big C and a bottle of 
>scotch has since eliminated them. And my personal email archive that went back 
>to '98 when I built my first linux machine using a 400 mhz k6 cpu. Put RedHat 
>5.0 on it.  And I was in hog heaven, I never owned a windows machine until I 
>needed one for the road after I retired in 2002 and became a consultant, going 
>around to other tv stations putting out engineering fires created by wannabe 
>engineers. The windows xp on it lasted about 2 weeks that it took me to find 
>out windows xp had no drivers for the radio in it, but mandrake did.
>
>Amanda keeps a database, so if something gets erased you need later, it could 
>be recovered as long as in my case 60 days later before the vtape has been 
>reused.
>
>One of the things my wrapper did was append that database to the end of that 
>vtape when amanda was finished from its nightly run, thereby making it 
>possible to do a bare metal recovery to the state that existed during the run. 
>Without that, you lost the most recent run because the database you backed up 
>was yesterdays.
>
>So AFAIAC, amanda was the king. Then amanda was handed over to Zmanda, who 
>eventually went bust and sold it to betsol, who has done zip for it in several 
>years.  Community support from other users is all thats left.
>Not the end of it of course, but somebody who actually cares needs to fork it 
>and become its new leader. 95% of the work on amanda has been driven by 
>changes in tar over the last decade+.
>
>> What are the use cases where Amanda still beats the pants down of
>> competitors like Borg or Bup?
>
>I know nothing about either of those. This thread ought to have input from 
>their users so people can make more informed decisions as to which is best for 
>their situation.
>> 
>> 
>>  Stefan
>Take care & stay well, Stefan, and other readers.
>> 
>> .
>
>Cheers, Gene Heskett.
>-- 
>"There are four boxes to be used in defense of liberty:
> soap, ballot, jury, and ammo. Please use in that order."

Re: Amanda (was: sata driver compataility Q)

[gene heskett]

On 9/19/23 08:59, Stefan Monnier wrote:

Compared to the setup required for amanda, that sounds very inviting. Amanda
has a very steep learning curve just because it is so versatile I'm still
waiting on stuff, so no more actual progress.


I used Amanda many years ago and was quite pleased with it, but I must
say I'm having a hard time imagining it in my current world where tapes
don't make much sense for backups.

Thats where vtapes come in, A vtape is nothing more than a directory on 
the backup medium, which for me was a BIG hard drive with in my case, 60 
suddirs, used as tapes. Each contained individual files identified as to 
backup level which was a way ti differentiate a full copy, or what had 
ben changed since the last full, or what had been changed since the last 
level 1, wash rinse repeat for ever deeper levels. And with or w/o 
compression. Executables generally aren't worth the time to compress. 
Ditto for a dir full if pictures or pdf's. They are not very 
compressible.  In the days of tapes, a buffer drive was used to build up 
each entry as a big file that was then copied to the tape w/o any 
shoe-shining of the tapedrive, saveibg the huge wear and tear of the 
tape if it had to stop and wait for data from the compressor. Then back 
uo a few feet, back forward to begin a fresh write at the end of the 
previous track. But since spinning rust is random access, amd so is the 
vtape, I don't think the anti-shoeshine has much if any advantage whrn 
using vtapes. With some filesystems it might reduce fragmentation but 
that was never a problem with ext4. I ran with that buffer drive for 
about 17 yeas, starting out with a 4 tape seagate dds4 tape drive but it 
was by far, the least dependable thing in that whole chain. I was then 
backing up 3 cnc machines and this ones predecessor, but the drive 
needed a months vacation in Oklahoma city about 2x a year for a new head 
drum that seagate would not sell me, a CET with extensive experience 
replacing even smaller, more precise and damn sure more expensive at 
$3500 a copy dvc-pro broadcast vcr heads.  So I tried vtapes, first on a 
220G drive but soon opted for a bigger one as they became available, and 
had just graduated to a pair of seagates first 2T's, both of which just 
disapeared off the sata buss in the middle of the night. the main drive 
for this machine and the amanda drive. They were about 2 weeks old. So I 
rebuilt this machine using a 500G Samsung SSD. I was out of the amanda 
business and lost everything with those 2 failures, whih upset me so 
much I never tried to warranty them. I was done with spinning rust.


Some of the loss was the only pix of my first wife who had a stroke and 
died in '68 at 34. Left me with 3 children to raise, but the big C and a 
bottle of scotch has since eliminated them. And my personal email 
archive that went back to '98 when I built my first linux machine using 
a 400 mhz k6 cpu. Put RedHat 5.0 on it.  And I was in hog heaven, I 
never owned a windows machine until I needed one for the road after I 
retired in 2002 and became a consultant, going around to other tv 
stations putting out engineering fires created by wannabe engineers. The 
windows xp on it lasted about 2 weeks that it took me to find out 
windows xp had no drivers for the radio in it, but mandrake did.


Amanda keeps a database, so if something gets erased you need later, it 
could be recovered as long as in my case 60 days later before the vtape 
has been reused.


One of the things my wrapper did was append that database to the end of 
that vtape when amanda was finished from its nightly run, thereby making 
it possible to do a bare metal recovery to the state that existed during 
the run. Without that, you lost the most recent run because the database 
you backed up was yesterdays.


So AFAIAC, amanda was the king. Then amanda was handed over to Zmanda, 
who eventually went bust and sold it to betsol, who has done zip for it 
in several years.  Community support from other users is all thats left.
Not the end of it of course, but somebody who actually cares needs to 
fork it and become its new leader. 95% of the work on amanda has been 
driven by changes in tar over the last decade+.



What are the use cases where Amanda still beats the pants down of
competitors like Borg or Bup?


I know nothing about either of those. This thread ought to have input 
from their users so people can make more informed decisions as to which 
is best for their situation.



 Stefan

Take care & stay well, Stefan, and other readers.


.


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, 1940)
If we desire respect for the law, we must first make the law respectable.
 - Louis D. Brandeis
Genes Web page 

Amanda (was: sata driver compataility Q)

[Stefan Monnier]

> Compared to the setup required for amanda, that sounds very inviting. Amanda
> has a very steep learning curve just because it is so versatile I'm still
> waiting on stuff, so no more actual progress.

I used Amanda many years ago and was quite pleased with it, but I must
say I'm having a hard time imagining it in my current world where tapes
don't make much sense for backups.

What are the use cases where Amanda still beats the pants down of
competitors like Borg or Bup?


Stefan

Re: sata driver compataility Q

[gene heskett]

On 9/18/23 07:57, Dan Ritter wrote:

gene heskett wrote:

[...]

It looks like the motherboard shares some PCIe and/or SATA lanes between
SATA ports and M.2 ports, so you must be careful with your choices.?? I
suggest installing an M.2 PCIe x4 SSD into slot M.2_1 and configuring it
for "PCIE mode", so that it works and all 6 SATA ports work.?? You will
want to use EUFI mode and GPT when installing Debian.



Based on this, and a full sized manual printout,  I've ordered a 2T WD
Black, supposedly a 2280 device. $100.

Question, when I put this in, what happens to the 32GB of dimms? How does
this fit into the architecture?  I assume this isn't volatile but is quick
storage.


The DIMM slots are different from the M.2 slots. The M.2 slots
are small PCIe interfaces; the installation procedure is to
insert the 2280 (22mm x 80mm) card at a slight upward angle,
then press it down and screw it in. It may ship with a glued-on
heat spreader or tiny radiator; if so, use it, don't peel it
off.

Note that it should appear as /dev/nvme0n1 or similar, rather
than /dev/sda. Partitions will be /dev/nvme0n1p1, p2...

The NVM bits stands for non-volatile memory; it's an SSD with a
different interface.


I propose to put this in as suggested, which should leave all 6 sata-III's
available, Install bookworm to it, w/o the current ectra controller get it
going, then put 3 of the 2T gigastones on sata1-2-3, use the bios to make a
raid5 of them and mount it as /home, prove it works with some throw away
stuff, then plug the existing raid10 controller & mount it as moi, then
format the raid5 again with gparted,


You should use mdadm rather than a BIOS RAID system -- better
recovery to other systems, more understandable error messages,
better support for fixing things that might go wrong.


Thanks, I will.


If the new drives are sda through sdf something like this is what
you want:

mdadm --create /dev/md/gsmoi5 -l raid5 -n 3 /dev/sda /dev/sdb /dev/sdc
mkfs.ext4 /dev/md/gsmoi5
mount /dev/md/gsmoi5 /home


and for your second set:

mdadm --create /dev/md/gsamanda5 -l raid5 -n 3 /dev/sdd /dev/sde /dev/sdf
mkfs.ext4 /dev/md/gsamanda5
mount /dev/md/gsamanda5 /amanda



Suggestions re other, more recent solutions will be accepted and studied.
Definitely must support backing up other machines of varying architectures
on my local network. In addition to a 4 pack of linux running wintel stuff,
there's the potential for 5 or so arms too. Gcodes for 3d printers are all
unrolled loops and bulky as can be.


The most flexible backup systems are the hardest to configure,
but nothing is much worse than amanda.

You might like borg. Borg is in Debian as 'borgbackup'.

In the other direction, using rsnapshot over ssh is relatively
simple and comes with the distinct advantage over both amanda
and borg that the backups are stored as normal files in a normal
filesystem, so recovery from an accidental deletion of a file or
directory is very straightforward.

Compared to the setup required for amanda, that sounds very inviting. 
Amanda has a very steep learning curve just because it is so versatile 
I'm still waiting on stuff, so no more actual progress.


Thanks Dan, tae care & stay well.

-dsr-
.


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, 1940)
If we desire respect for the law, we must first make the law respectable.
 - Louis D. Brandeis
Genes Web page 

Re: sata driver compataility Q

[David Christensen]

On 9/17/23 18:17, gene heskett wrote:

On 9/17/23 17:52, David Christensen wrote:

On 9/17/23 03:26, gene heskett wrote:

On 9/16/23 19:46, David Christensen wrote:

On 9/15/23 19:37, gene heskett wrote:

On 9/15/23 20:12, David Christensen wrote:

On 9/15/23 15:04, gene heskett wrote:

On 9/15/23 17:35, David Christensen wrote:

On 9/15/23 12:28, gene heskett wrote:

I've just ordered some stuff to rebuild or expand my Raid setup.
This 16 port sata-III pci-e card:
<https://www.amazon.com/dp/B09L184W57?smid=A2H818KAC5I4D1_=chk_typ_imgToDp=1>
along with a bigger drive cage, cables and such and some 
gigastone 2T drives to make a raid big enough to run amanda. 
And maybe put a new card in front of my 2T /home raid10.



... Asus PRIME Z370-A II motherboard ...



... i5-9600K CPU @ 3.70GHz ...


A fresh install of  Debian stable or old-stable should solve the 
storage I/O stuttering problems you are experiencing.


It looks like the motherboard shares some PCIe and/or SATA lanes 
between SATA ports and M.2 ports, so you must be careful with your 
choices.  I suggest installing an M.2 PCIe x4 SSD into slot M.2_1 and 
configuring it for "PCIE mode", so that it works and all 6 SATA ports 
work.  You will want to use EUFI mode and GPT when installing Debian.


Based on this, and a full sized manual printout,  I've ordered a 2T WD 
Black, supposedly a 2280 device. $100.



This one?

https://www.amazon.com/dp/B09QV5KJHV


Question, when I put this in, what happens to the 32GB of dimms?  How 
does this fit into the architecture?  I assume this isn't volatile but 
is quick storage.



You will still have 32GB of memory.  The WD Black is a fast SSD.  The 
crux will be configuring your motherboard firmware Setup program so that 
d-i can see the WD Black during installation and so that the new Debian 
installation can boot and run.



I propose to put this in as suggested, which should leave all 6 
sata-III's available, Install bookworm to it, w/o the current ectra 
controller get it going, then put 3 of the 2T gigastones on sata1-2-3, 
use the bios to make a raid5 of them and mount it as /home, prove it 
works with some throw away stuff, then plug the existing raid10 
controller & mount it as moi, then format the raid5 again with gparted,
and turn mc loose copying /moi to /home to get my working data back. 
Then 3 more 2T gigastones on the last 3 mobo sata ports, make another 
raid5 out of those mounted as amandatapes.. Unforch, I had a wrapper 
around amanda that it took me 5 years to fine tune but I've no idea if a 
backup copy exists anyplace it this midden heap. Amanda, as it exists, 
if you start a recovery to bare metal, can only restore to yesterday, my 
wrapper was a special deal that grabbed the database fom the just 
finished backup and put that into the vtape, uncompressed which if that 
was untared to the bare metal gave anmanda the data for recovery that 
would bring the system back to this mornings state. It also cleaned out 
the database of any links that referenced vtapes that were recycled and 
re-used.



Most hardware- and hybrid hardware/ software RAID solutions expect 
Windows -- e.g. the manufacturer provides a Windows bundle with device 
driver, CLI, GUI, etc..  Looking at the Asus PRIME Z370-A II Driver & 
Tools page, I see various Windows packages related to storage, but 
nothing for Linux.  Unless you can find suitable Debian packages, I 
would advise against motherboard RAID.



Debian supports software RAID via md, LVM, and btrfs.  I suggest that 
you use one of those.



ZFS is another possibility, but the learning curve is non-trivial.


So while I'm familiar with amanda, its been sold to an outfit that 
doesn't care, so its getting long in the tooth with only user support.


Suggestions re other, more recent solutions will be accepted and 
studied.  Definitely must support backing up other machines of varying 
architectures on my local network. In addition to a 4 pack of linux 
running wintel stuff, there's the potential for 5 or so arms too. Gcodes 
for 3d printers are all unrolled loops and bulky as can be.



I suggest starting with the WD Black and the new Debian installation.  A 
fresh install on a new device will simplify re-arranging the rest of 
your disks later.  The challenge will be deciding what data to put on it 
after Debian boot, swap, and root; and if and how to subdivide the space.




Thank you David, take care and stay well.


Likewise.  :-)


David

Re: sata driver compataility Q

[gene heskett]

On 9/17/23 17:52, David Christensen wrote:

On 9/17/23 03:26, gene heskett wrote:

On 9/16/23 19:46, David Christensen wrote:

On 9/15/23 19:37, gene heskett wrote:

On 9/15/23 20:12, David Christensen wrote:

On 9/15/23 15:04, gene heskett wrote:

On 9/15/23 17:35, David Christensen wrote:

On 9/15/23 12:28, gene heskett wrote:

I've just ordered some stuff to rebuild or expand my Raid setup.
This 16 port sata-III pci-e card:

along with a bigger drive cage, cables and such and some 
gigastone 2T drives to make a raid big enough to run amanda. And 
maybe put a new card in front of my 2T /home raid10.


Searching the mailing list archive, it appears that you have an Asus 
PRIME Z370-A II motherboard (?):


https://www.asus.com/motherboards-components/motherboards/prime/prime-z370-a-ii/

And, an Intel Core i5 processor (?).  Which model?


cpuinfo can't copy/paste, 6 core, i5-9600K CPU @ 3.70GHz ...



Okay.  I do not overclock, but the "K" suffix processor is usually the 
fastest OOTB in a given series.



How many GB of OS and apps do you have?  Home directory?  Bulk data? 
Amanda backups?  VM's?  Other?



Knowing how much and what kind of live, backup, and whatever data you 
have would help us make better suggestions for storage.  Similarly, your 
drives, HBA's, chassis, and chassis mods (notably drive bays).



While researching this thread, I came across an HBA that may interest 
both of us:


 https://www.amazon.com/dp/B09L3GLCL9

4x the bandwidth (3.94 GB/s), 8 more SATA ports (24 total), and $42.31
higher price ($117.30).  I would prefer this card for the bandwidth
alone, and I never know when I might need those extra ports to
temporarily connect a bunch of disks from other machines.


There is an Amazon review that states the IO CREST is PCIe 3.0 x2 
electrically.  If correct, the bandwidth is 1.97 GB/s, which should be 
sufficient to saturate about a dozen HDD's.



The Amazon page for the IO CREST provides a part number:

     SI-PEX40169


STFW for the part number leads to a Syba HBA:

 
https://www.sybausa.com/index.php?route=product/product_id=1095


     Uses 5 JMB575 ( 8 SATA Ports Per Slim SAS) connected to a JMB582 via
     Port Multiplier Mode


STFW for the chips:

* JMB582

     https://www.jmicron.com/products/list/15

     PCIe Gen3 x1 to Dual SATA 6Gb/s

* JMB575

     https://www.jmicron.com/products/list/16

     1 to 5 ports SATA 6Gb/s
     Port Multiplier / Port Selector


STFW it is hard to find good technical information, infer the 
architecture, or reason about the performance of the HBA.



I have some Syba PCIe 1.0 x1 to 2 @ SATA II HBA's on the shelf.  They 
have worked with Windows, Debian, and/or FreeBSD for many years.



Thats nice but only an x4 for $117, how about an x16, which I have an 
empty slot for a full length X16 at $160: > 
https://www.amazon.com/dp/B09K5GLJ8D



The Amazon page for the BEYIMEI HBA states:

     Use chipset 6 * ASM1064 + ASM1812 * 1 main control chip


STFW for those chips:

* ASM1064:

     https://www.asmedia.com.tw/product/A58yQC9Sp5qg6TrF/58dYQ8bxZ4UR9wG5

     ASM1064, a SATA host controller(AHCI) with upstream PCIe Gen3 x1 and
     downstream four SATA Gen3 ports. It’s a low latency, low cost and
     low power AHCI controller. With four SATA ports and cascaded port
     multipliers, ASM1064 can enable users to build up various high speed
     IO systems, including server, high capacity system storage or
     surveillance platforms.

* ASM1812:

     https://www.asmedia.com.tw/product/1e2yQ48sx5HT2pUF/b7FyQBCxz2URbzg0

     ASMedia PCIe product ASM1812, a low latency, low cost and low power
     12 lane , maximum 6 downstream ports packet switch. With upstream
     PCIe Gen2x4 bandwidth, ASM1812 can enable users to build up various
     high speed IO systems, including server, system storage or
     communication platforms.


Again, it is hard to find technical information, infer the architecture, 
or reason about performance of the HBA.




Or is the jmicron x4 better supported?



I do not know.  One option is to STFW for release notes, bug reports, 
reviews, etc..  After that, I expect it boils down to "buy it, try it; 
return if not satisfied".



Regardless of what you do with HBA's, I would connect the six SSD's 
to the six motherboard SATA III ports.


Which would tie them up. Is that the faster solution, moving the 
opticals to a different card? 



My optical drives are SATA I (1.5 Gbps).  I connect them to the slowest 
SATA ports in my computers.  I use the motherboard SATA III ports for my 
fastest SATA drives under the heaviest workloads -- SSD system drives 
and cache/ ZIL/ dedup vdev's.




The question then is can the bios see them to boot from them?



That depends upon the SATA port and the drive.  If you must, you can 
make a temporary connection while booting and running an optical disc.



I did grab what I think is a newer bios 

Re: sata driver compataility Q

[David Christensen]

On 9/17/23 03:26, gene heskett wrote:

On 9/16/23 19:46, David Christensen wrote:

On 9/15/23 19:37, gene heskett wrote:

On 9/15/23 20:12, David Christensen wrote:

On 9/15/23 15:04, gene heskett wrote:

On 9/15/23 17:35, David Christensen wrote:

On 9/15/23 12:28, gene heskett wrote:

I've just ordered some stuff to rebuild or expand my Raid setup.
This 16 port sata-III pci-e card:

along with a bigger drive cage, cables and such and some 
gigastone 2T drives to make a raid big enough to run amanda. And 
maybe put a new card in front of my 2T /home raid10.


Searching the mailing list archive, it appears that you have an Asus 
PRIME Z370-A II motherboard (?):


https://www.asus.com/motherboards-components/motherboards/prime/prime-z370-a-ii/

And, an Intel Core i5 processor (?).  Which model?


cpuinfo can't copy/paste, 6 core, i5-9600K CPU @ 3.70GHz ...



Okay.  I do not overclock, but the "K" suffix processor is usually the 
fastest OOTB in a given series.



How many GB of OS and apps do you have?  Home directory?  Bulk data? 
Amanda backups?  VM's?  Other?



Knowing how much and what kind of live, backup, and whatever data you 
have would help us make better suggestions for storage.  Similarly, your 
drives, HBA's, chassis, and chassis mods (notably drive bays).



While researching this thread, I came across an HBA that may interest 
both of us:


 https://www.amazon.com/dp/B09L3GLCL9

4x the bandwidth (3.94 GB/s), 8 more SATA ports (24 total), and $42.31
higher price ($117.30).  I would prefer this card for the bandwidth
alone, and I never know when I might need those extra ports to
temporarily connect a bunch of disks from other machines.


There is an Amazon review that states the IO CREST is PCIe 3.0 x2 
electrically.  If correct, the bandwidth is 1.97 GB/s, which should be 
sufficient to saturate about a dozen HDD's.



The Amazon page for the IO CREST provides a part number:

SI-PEX40169


STFW for the part number leads to a Syba HBA:

https://www.sybausa.com/index.php?route=product/product_id=1095

Uses 5 JMB575 ( 8 SATA Ports Per Slim SAS) connected to a JMB582 via
Port Multiplier Mode


STFW for the chips:

* JMB582

https://www.jmicron.com/products/list/15

PCIe Gen3 x1 to Dual SATA 6Gb/s

* JMB575

https://www.jmicron.com/products/list/16

1 to 5 ports SATA 6Gb/s
Port Multiplier / Port Selector


STFW it is hard to find good technical information, infer the 
architecture, or reason about the performance of the HBA.



I have some Syba PCIe 1.0 x1 to 2 @ SATA II HBA's on the shelf.  They 
have worked with Windows, Debian, and/or FreeBSD for many years.



Thats nice but only an x4 for $117, how about an x16, which I have an 
empty slot for a full length X16 at $160: > https://www.amazon.com/dp/B09K5GLJ8D



The Amazon page for the BEYIMEI HBA states:

Use chipset 6 * ASM1064 + ASM1812 * 1 main control chip


STFW for those chips:

* ASM1064:

https://www.asmedia.com.tw/product/A58yQC9Sp5qg6TrF/58dYQ8bxZ4UR9wG5

ASM1064, a SATA host controller(AHCI) with upstream PCIe Gen3 x1 and
downstream four SATA Gen3 ports. It’s a low latency, low cost and
low power AHCI controller. With four SATA ports and cascaded port
multipliers, ASM1064 can enable users to build up various high speed
IO systems, including server, high capacity system storage or
surveillance platforms.

* ASM1812:

https://www.asmedia.com.tw/product/1e2yQ48sx5HT2pUF/b7FyQBCxz2URbzg0

ASMedia PCIe product ASM1812, a low latency, low cost and low power
12 lane , maximum 6 downstream ports packet switch. With upstream
PCIe Gen2x4 bandwidth, ASM1812 can enable users to build up various
high speed IO systems, including server, system storage or
communication platforms.


Again, it is hard to find technical information, infer the architecture, 
or reason about performance of the HBA.




Or is the jmicron x4 better supported?



I do not know.  One option is to STFW for release notes, bug reports, 
reviews, etc..  After that, I expect it boils down to "buy it, try it; 
return if not satisfied".



Regardless of what you do with HBA's, I would connect the six SSD's to 
the six motherboard SATA III ports.


Which would tie them up. Is that the faster solution, moving the 
opticals to a different card? 



My optical drives are SATA I (1.5 Gbps).  I connect them to the slowest 
SATA ports in my computers.  I use the motherboard SATA III ports for my 
fastest SATA drives under the heaviest workloads -- SSD system drives 
and cache/ ZIL/ dedup vdev's.



The question then is can the bios see them 
to boot from them?



That depends upon the SATA port and the drive.  If you must, you can 
make a temporary connection while booting and running an optical disc.



I did grab what I think is a newer bios yesterday but haven't tried to 
install it yet.

  

Re: sata driver compataility Q

[Andrew M.A. Cater]

On Sun, Sep 17, 2023 at 06:26:49AM -0400, gene heskett wrote:
> On 9/16/23 19:46, David Christensen wrote:
> > On 9/15/23 19:37, gene heskett wrote:
> > > On 9/15/23 20:12, David Christensen wrote:
> > > > On 9/15/23 15:04, gene heskett wrote:
> > > > > On 9/15/23 17:35, David Christensen wrote:
> > > > > > On 9/15/23 12:28, gene heskett wrote:
> > > > > > > I've just ordered some stuff to rebuild or expand my Raid setup.
> > > > > > > This 16 port sata-III pci-e card:
> > > > > > > 
> > > > > > > along with a bigger drive cage, cables and such and
> > > > > > > some gigastone 2T drives to make a raid big enough
> > > > > > > to run amanda. And maybe put a new card in front of
> > > > > > > my 2T /home raid10.
> > > > > > 
> > > > > > Is everything going into one chassis?  Have you
> > > > > > considered an external drive chassis?
> > > > > 
> > > 
> > > Call me cheap, my choice is diy assembly, SS stampings you put
> > > together, all drive brackets and a dozen cables and a 5 drive power
> > > splitter, $30.
> > > Less than 5% of the price of that nice looking box.
> > 
> > 
> > Fair enough.
> > 
> > 
> > Searching the mailing list archive, it appears that you have an Asus
> > PRIME Z370-A II motherboard (?):
> > 
> > 
> > https://www.asus.com/motherboards-components/motherboards/prime/prime-z370-a-ii/
> > 
> > 
> > And, an Intel Core i5 processor (?).  Which model?
> 
> cpuinfo can't copy/paste, 6 core, i5-9600K CPU @ 3.70GHz but apparently very
> pushable, I ran one cycle of memtester pointed at 16G's and saw it above
> 4300 mhz a couple times, and that core hit 51C, which is 20C hotter than its
> ever run b4. With my intermittent load it often loafs at 800 mhz & 29C, 33C
> when OpenSCAD is munching on something I've designed.
> 
> > How many GB of OS and apps do you have?  Home directory?  Bulk data?
> > Amanda backups?  VM's?  Other?
> > 
> > 
> > 
> > 4x the bandwidth (3.94 GB/s), 8 more SATA ports (24 total), and $42.31
> > higher price ($117.30).  I would prefer this card for the bandwidth
> > alone, and I never know when I might need those extra ports to
> > temporarily connect a bunch of disks from other machines.
> > 
> > 
> > If your goal is maximum capacity at minimum cost, HDD's have larger
> > capacity, lower bandwidth, and lower cost per TB than SSD's -- per bay,
> > per drive, and per port.  Port multiplication makes sense with HDD's.
> > 
> > 
> > For file server and backup server roles, I use ZFS with HDD primary
> > storage devices and full bandwidth HBA.  Read performance, write
> > performance, and capacity utilization can be improved with ZFS
> > compression, read performance can be improved with an SSD cache vdev
> > (virtual device; e.g. partition), and write performance can be improved
> > with SSD intent log vdev (mirror of partitions).  For the backup server
> > role, capacity utilization can be improved with ZFS deduplication and an
> > SSD dedup vdev (mirror of partitions).
> > 
> > 
> > Regardless of what you do with HBA's, I would connect the six SSD's to
> > the six motherboard SATA III ports.
> > 
> Which would tie them up. Is that the faster solution, moving the opticals to
> a different card? The question then is can the bios see them to boot from
> them?
> 
> I did grab what I think is a newer bios yesterday but haven't tried to
> install it yet.
>  mt86plus_6.20_64.grub.iso.zip
> > 
> > A fresh install of  Debian stable or old-stable should solve the storage
> > I/O stuttering problems you are experiencing.  (That motherboard has
> > dual M.2 ports.  Installing Debian onto an M.2 PCIE 3.0 x4 SSD would be
> > very nice.)
> > 

See below: maybe get a new machine to do this on ...

> I know zip about the new m2 stuff. Link to good info appreciated. I didn't
> use it originally because I couldn't find the m2 sockets but probably didn't
> look too hard as I was more concerned with getting another system built
> after a usb socket on the previous occupant of that space caught fire and
> tried to burn the place down.
> 

Given what you have: m2 is probably faster than anything else you've got.
Chuck a terabyte drive in and forget about running Debian anywhere else
for your OS.

Any add in card is likely to be slower than your MB sata ports which are
relatively well connected / you will end up with contention between a 
relatively fast card and the motherboard with a bottleneck somewhere.

Unless you buy the very best server grade cards - and even then - cheap
cards are doing everything in software so you're just running an expensive
bunch of disks and doing software RAID anyway - at which point you might as
well be using mdadm. ZFS - a whole new learning curve for you, a bunch of
tuning you may or may not want to do - and another thing for you to blame
in due course, perhaps..

You have a fascination for buying drives and kludging things together -
you might honestly be better getting yourself a 

Re: sata driver compataility Q

[gene heskett]

On 9/16/23 19:46, David Christensen wrote:

On 9/15/23 19:37, gene heskett wrote:

On 9/15/23 20:12, David Christensen wrote:

On 9/15/23 15:04, gene heskett wrote:

On 9/15/23 17:35, David Christensen wrote:

On 9/15/23 12:28, gene heskett wrote:

I've just ordered some stuff to rebuild or expand my Raid setup.
This 16 port sata-III pci-e card:

along with a bigger drive cage, cables and such and some gigastone 
2T drives to make a raid big enough to run amanda. And maybe put a 
new card in front of my 2T /home raid10.


Is everything going into one chassis?  Have you considered an 
external drive chassis?


Got one of those coming too. Small possibility it will fit it the 
bottom of this huge and old tiger direct tower.  Because the 
radiator for the 6 core i5 is too tall, it hasn't had a side cover 
on it in years. If not, theres a 3 3.5" cage at the bottom of the 
stack I can stuff  with 2 SSD's per bay slot. A hidey place, front 
cover is solid, but cooling might be a problem. If worse comes to 
worse I could shoe-goo a 120x15 to the side of the cage.



This is what I meant:

 https://www.pc-pitstop.com/16-bay-25inch-sas-sata-jbod-tower


Call me cheap, my choice is diy assembly, SS stampings you put 
together, all drive brackets and a dozen cables and a 5 drive power 
splitter, $30.

Less than 5% of the price of that nice looking box.



Fair enough.


Searching the mailing list archive, it appears that you have an Asus 
PRIME Z370-A II motherboard (?):



https://www.asus.com/motherboards-components/motherboards/prime/prime-z370-a-ii/


And, an Intel Core i5 processor (?).  Which model?


cpuinfo can't copy/paste, 6 core, i5-9600K CPU @ 3.70GHz but apparently 
very pushable, I ran one cycle of memtester pointed at 16G's and saw it 
above 4300 mhz a couple times, and that core hit 51C, which is 20C 
hotter than its ever run b4. With my intermittent load it often loafs at 
800 mhz & 29C, 33C when OpenSCAD is munching on something I've designed.


How many GB of OS and apps do you have?  Home directory?  Bulk data? 
Amanda backups?  VM's?  Other?



While researching this thread, I came across an HBA that may interest 
both of us:



     https://www.amazon.com/dp/B09L3GLCL9

Thats nice but only an x4 for $117, how about an x16, which I have an 
empty slot for a full length X16 at $160:



Or is the jmicron x4 better supported?


4x the bandwidth (3.94 GB/s), 8 more SATA ports (24 total), and $42.31 
higher price ($117.30).  I would prefer this card for the bandwidth 
alone, and I never know when I might need those extra ports to 
temporarily connect a bunch of disks from other machines.



If your goal is maximum capacity at minimum cost, HDD's have larger 
capacity, lower bandwidth, and lower cost per TB than SSD's -- per bay, 
per drive, and per port.  Port multiplication makes sense with HDD's.



For file server and backup server roles, I use ZFS with HDD primary 
storage devices and full bandwidth HBA.  Read performance, write 
performance, and capacity utilization can be improved with ZFS 
compression, read performance can be improved with an SSD cache vdev 
(virtual device; e.g. partition), and write performance can be improved 
with SSD intent log vdev (mirror of partitions).  For the backup server 
role, capacity utilization can be improved with ZFS deduplication and an 
SSD dedup vdev (mirror of partitions).



Regardless of what you do with HBA's, I would connect the six SSD's to 
the six motherboard SATA III ports.


Which would tie them up. Is that the faster solution, moving the 
opticals to a different card? The question then is can the bios see them 
to boot from them?


I did grab what I think is a newer bios yesterday but haven't tried to 
install it yet.

 mt86plus_6.20_64.grub.iso.zip


A fresh install of  Debian stable or old-stable should solve the storage 
I/O stuttering problems you are experiencing.  (That motherboard has 
dual M.2 ports.  Installing Debian onto an M.2 PCIE 3.0 x4 SSD would be 
very nice.)


I know zip about the new m2 stuff. Link to good info appreciated. I 
didn't use it originally because I couldn't find the m2 sockets but 
probably didn't look too hard as I was more concerned with getting 
another system built after a usb socket on the previous occupant of that 
space caught fire and tried to burn the place down.


Can you find a link showing 

Re: sata driver compataility Q

[David Christensen]

On 9/15/23 19:37, gene heskett wrote:

On 9/15/23 20:12, David Christensen wrote:

On 9/15/23 15:04, gene heskett wrote:

On 9/15/23 17:35, David Christensen wrote:

On 9/15/23 12:28, gene heskett wrote:

I've just ordered some stuff to rebuild or expand my Raid setup.
This 16 port sata-III pci-e card:

along with a bigger drive cage, cables and such and some gigastone 
2T drives to make a raid big enough to run amanda. And maybe put a 
new card in front of my 2T /home raid10.


Is everything going into one chassis?  Have you considered an 
external drive chassis?


Got one of those coming too. Small possibility it will fit it the 
bottom of this huge and old tiger direct tower.  Because the radiator 
for the 6 core i5 is too tall, it hasn't had a side cover on it in 
years. If not, theres a 3 3.5" cage at the bottom of the stack I can 
stuff  with 2 SSD's per bay slot. A hidey place, front cover is 
solid, but cooling might be a problem. If worse comes to worse I 
could shoe-goo a 120x15 to the side of the cage.



This is what I meant:

 https://www.pc-pitstop.com/16-bay-25inch-sas-sata-jbod-tower


Call me cheap, my choice is diy assembly, SS stampings you put together, 
all drive brackets and a dozen cables and a 5 drive power splitter, $30.

Less than 5% of the price of that nice looking box.



Fair enough.


Searching the mailing list archive, it appears that you have an Asus 
PRIME Z370-A II motherboard (?):



https://www.asus.com/motherboards-components/motherboards/prime/prime-z370-a-ii/


And, an Intel Core i5 processor (?).  Which model?


How many GB of OS and apps do you have?  Home directory?  Bulk data? 
Amanda backups?  VM's?  Other?



While researching this thread, I came across an HBA that may interest 
both of us:



https://www.amazon.com/dp/B09L3GLCL9


4x the bandwidth (3.94 GB/s), 8 more SATA ports (24 total), and $42.31 
higher price ($117.30).  I would prefer this card for the bandwidth 
alone, and I never know when I might need those extra ports to 
temporarily connect a bunch of disks from other machines.



If your goal is maximum capacity at minimum cost, HDD's have larger 
capacity, lower bandwidth, and lower cost per TB than SSD's -- per bay, 
per drive, and per port.  Port multiplication makes sense with HDD's.



For file server and backup server roles, I use ZFS with HDD primary 
storage devices and full bandwidth HBA.  Read performance, write 
performance, and capacity utilization can be improved with ZFS 
compression, read performance can be improved with an SSD cache vdev 
(virtual device; e.g. partition), and write performance can be improved 
with SSD intent log vdev (mirror of partitions).  For the backup server 
role, capacity utilization can be improved with ZFS deduplication and an 
SSD dedup vdev (mirror of partitions).



Regardless of what you do with HBA's, I would connect the six SSD's to 
the six motherboard SATA III ports.



A fresh install of  Debian stable or old-stable should solve the storage 
I/O stuttering problems you are experiencing.  (That motherboard has 
dual M.2 ports.  Installing Debian onto an M.2 PCIE 3.0 x4 SSD would be 
very nice.)



David

Re: sata driver compataility Q

[songbird]

gene heskett wrote:
> On 9/16/23 06:07, songbird wrote:
>> gene heskett wrote:
>> ...
>>> This setup worked instantly under buster and bullseye, but takes from 30
>>> secs to 5 minutes to open a write requestor window asking where to put
>>> the download I clicked on under bookworrm.
>> 
>>trace the first part of the process and see what is
>> taking so long.
>
> I'd love to, but how do you trace a mouse click?  All the 
> "alphabet-trace" utils I know are cli only. Probably my fault, but...

  replace the /usr/bin/ by a script which traces 
the  in question...  or if it is X related there
are probably ways of tracing X.

  it might be a desktop, greeter or a window manager 
binary of some type but you should eventually be able 
to figure out what is going on.


  songbird

Re: sata driver compataility Q

[gene heskett]

On 9/16/23 06:07, songbird wrote:

gene heskett wrote:
...

This setup worked instantly under buster and bullseye, but takes from 30
secs to 5 minutes to open a write requestor window asking where to put
the download I clicked on under bookworrm.


   trace the first part of the process and see what is
taking so long.


I'd love to, but how do you trace a mouse click?  All the 
"alphabet-trace" utils I know are cli only. Probably my fault, but...



   songbird

.


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, 1940)
If we desire respect for the law, we must first make the law respectable.
 - Louis D. Brandeis
Genes Web page 

Re: sata driver compataility Q

[songbird]

gene heskett wrote:
...
> This setup worked instantly under buster and bullseye, but takes from 30 
> secs to 5 minutes to open a write requestor window asking where to put 
> the download I clicked on under bookworrm.

  trace the first part of the process and see what is 
taking so long.


  songbird

Re: sata driver compataility Q

[gene heskett]

On 9/15/23 20:12, David Christensen wrote:

On 9/15/23 15:04, gene heskett wrote:

On 9/15/23 17:35, David Christensen wrote:

On 9/15/23 12:28, gene heskett wrote:

Greetings all;

I've just ordered some stuff to rebuild or expand my Raid setup.
This 16 port sata-III pci-e card:

along with a bigger drive cage, cables and such and some gigastone 
2T drives to make a raid big enough to run amanda. And maybe put a 
new card in front of my 2T /home raid10.


Is everything going into one chassis?  Have you considered an 
external drive chassis?


Got one of those coming too. Small possibility it will fit it the 
bottom of this huge and old tiger direct tower.  Because the radiator 
for the 6 core i5 is too tall, it hasn't had a side cover on it in 
years. If not, theres a 3 3.5" cage at the bottom of the stack I can 
stuff  with 2 SSD's per bay slot. A hidey place, front cover is solid, 
but cooling might be a problem. If worse comes to worse I could 
shoe-goo a 120x15 to the side of the cage.



This is what I meant:

     https://www.pc-pitstop.com/16-bay-25inch-sas-sata-jbod-tower


David
Call me cheap, my choice is diy assembly, SS stampings you put together, 
all drive brackets and a dozen cables and a 5 drive power splitter, $30.

Less than 5% of the price of that nice looking box.


.


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, 1940)
If we desire respect for the law, we must first make the law respectable.
 - Louis D. Brandeis
Genes Web page 

Re: sata driver compataility Q

[David Christensen]

On 9/15/23 15:04, gene heskett wrote:

On 9/15/23 17:35, David Christensen wrote:

On 9/15/23 12:28, gene heskett wrote:

Greetings all;

I've just ordered some stuff to rebuild or expand my Raid setup.
This 16 port sata-III pci-e card:

along with a bigger drive cage, cables and such and some gigastone 2T 
drives to make a raid big enough to run amanda. And maybe put a new 
card in front of my 2T /home raid10.


Is everything going into one chassis?  Have you considered an external 
drive chassis?


Got one of those coming too. Small possibility it will fit it the bottom 
of this huge and old tiger direct tower.  Because the radiator for the 6 
core i5 is too tall, it hasn't had a side cover on it in years. If not, 
theres a 3 3.5" cage at the bottom of the stack I can stuff  with 2 
SSD's per bay slot. A hidey place, front cover is solid, but cooling 
might be a problem. If worse comes to worse I could shoe-goo a 120x15 to 
the side of the cage.



This is what I meant:

https://www.pc-pitstop.com/16-bay-25inch-sas-sata-jbod-tower


David

Re: sata driver compataility Q

[gene heskett]

On 9/15/23 17:56, Andy Smith wrote:

Hello,

On Fri, Sep 15, 2023 at 05:35:40PM -0400, gene heskett wrote:

This setup worked instantly under buster and bullseye, but takes from 30
secs to 5 minutes to open a write requestor window asking where to put the
download I clicked on under bookworrm.


I think you should work out why that happens before spending a lot
of money on new hardware. It doesn't seem at all likely to me that
your existing hardware is at fault. Buying new hardware risks
experiencing the same thing with still no idea why.

Thanks,
Andy

I won't argue on that point Andy, but it now been several months asking 
the same question from different angles and not getting a single helpful 
reply. I know my reputation is bad because I often use the box I'm 
supposed to stay inside of, as kindling to start the next campfire. I 
don't "stay inside that famous box" unless there is someone outside 
shooting at me, it which case I might shoot back. Computers can do 
anything you can write an interface to tickle the hardware for.


Now if someone can have me check something that might be aglay, my 
fingers to do that check are at your disposal. It eventually works, 100% 
of the time. But why the 30 sec to 5 minute delay before it allows me to 
do what I asked?


Frustrated is the PC word, but that is certainly not how I describe it 
in my lonelyness. There is not anyone here to hear me rant except me, 
which is probably a "good thing".


Thanks, take care and stay well Andy.

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, 1940)
If we desire respect for the law, we must first make the law respectable.
 - Louis D. Brandeis
Genes Web page 

Re: sata driver compataility Q

[gene heskett]

On 9/15/23 17:35, David Christensen wrote:

On 9/15/23 12:28, gene heskett wrote:

Greetings all;

I've just ordered some stuff to rebuild or expand my Raid setup.
This 16 port sata-III pci-e card:

along with a bigger drive cage, cables and such and some gigastone 2T 
drives to make a raid big enough to run amanda. And maybe put a new 
card in front of my 2T /home raid10.


The card claims linux compatibility.

Can anyone advise me on the gotcha's of such a 16 port beast?

Thanks all.

Cheers, Gene Heskett.



Searching Amazon for "gigastone 2T", I see:


https://www.amazon.com/Gigastone-Internal-Compatible-Desktop-Laptop/dp/B0BN5978X1

     540 MB per second


PCIe 3.0 x1 is rated for 985 MB/s.

     https://en.wikipedia.org/wiki/Pcie


So, the PCIe 3.0 x1 connector is going to be a bottleneck when accessing 
more than one SSD.



I suggest that you pick an HBA with a wider PCIe connector -- PCIe 3.0 
x8 (7.88 GB/s) is a reasonable match for sixteen SSD's (8.64 GB/s). PCIe 
3.0 x16 would eliminate the PCIe bottleneck.



Is everything going into one chassis?  Have you considered an external 
drive chassis?
Got one of those coming too. Small possibility it will fit it the bottom 
of this huge and old tiger direct tower.  Because the radiator for the 6 
core i5 is too tall, it hasn't had a side cover on it in years. If not, 
theres a 3 3.5" cage at the bottom of the stack I can stuff  with 2 
SSD's per bay slot. A hidey place, front cover is solid, but cooling 
might be a problem. If worse comes to worse I could shoe-goo a 120x15 to 
the side of the cage.



Make sure your power supply(s) are adequate to the task.


400 watter, presently runs dead cold. 5V line rated at 16A, probably not 
using half that now.


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, 1940)
If we desire respect for the law, we must first make the law respectable.
 - Louis D. Brandeis
Genes Web page 

Re: sata driver compataility Q

[Andy Smith]

Hello,

On Fri, Sep 15, 2023 at 05:35:40PM -0400, gene heskett wrote:
> This setup worked instantly under buster and bullseye, but takes from 30
> secs to 5 minutes to open a write requestor window asking where to put the
> download I clicked on under bookworrm.

I think you should work out why that happens before spending a lot
of money on new hardware. It doesn't seem at all likely to me that
your existing hardware is at fault. Buying new hardware risks
experiencing the same thing with still no idea why.

Thanks,
Andy

-- 
https://bitfolk.com/ -- No-nonsense VPS hosting

Re: sata driver compataility Q

[gene heskett]

On 9/15/23 15:56, Dan Ritter wrote:

gene heskett wrote:

Greetings all;

I've just ordered some stuff to rebuild or expand my Raid setup.
This 16 port sata-III pci-e card:

along with a bigger drive cage, cables and such and some gigastone 2T drives
to make a raid big enough to run amanda. And maybe put a new card in front
of my 2T /home raid10.

The card claims linux compatibility.

Can anyone advise me on the gotcha's of such a 16 port beast?


What you actually have there is 1 SATA-3 controller which should
be able to support 4 disks, and 4 port multipliers to support
16.

And I forgot to ask, which would be faster, 4 drives on port1 1-4, or 4 
drives on ports 1-5-9-13?



Effectively, every group of four disks is competing against
themselves. So performance is going to be mediocre.

It should work, though, as a giant dumping ground.

But why are you buying 16 x 2TB disks, if not for performance?

I've never heard of Gigastone and can offer no assessment of
them.

-dsr-
.


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, 1940)
If we desire respect for the law, we must first make the law respectable.
 - Louis D. Brandeis
Genes Web page 

Re: sata driver compataility Q

[gene heskett]

On 9/15/23 15:56, Dan Ritter wrote:

gene heskett wrote:

Greetings all;

I've just ordered some stuff to rebuild or expand my Raid setup.
This 16 port sata-III pci-e card:

along with a bigger drive cage, cables and such and some gigastone 2T drives
to make a raid big enough to run amanda. And maybe put a new card in front
of my 2T /home raid10.

The card claims linux compatibility.

Can anyone advise me on the gotcha's of such a 16 port beast?


What you actually have there is 1 SATA-3 controller which should
be able to support 4 disks, and 4 port multipliers to support
16.

Effectively, every group of four disks is competing against
themselves. So performance is going to be mediocre.

Since the pci-e plug is the narrow one I suspected as much, the data 
path is too narrow, This asus mobo has 6 ports but I'd assume the mobo 
ports would be wider, perhaps even x4 but it is also not stated,
The existing 4 1T per drive raid10 is on its own x1 based 6 port 
controller. The other 2 ports are not used at present,


This setup worked instantly under buster and bullseye, but takes from 30 
secs to 5 minutes to open a write requestor window asking where to put 
the download I clicked on under bookworrm.  And just as often as I've 
mentioned it the subject of the reply if any, is changed w/o changing 
the subject line, to ignore it.




It should work, though, as a giant dumping ground.

Such as an amanda backup raid, and running it the wee hours, I could 
care less how long it takes as long as it is done by 05:30 or 06:00. If 
I can rearrange the usb breakouts, I can uncover another pci-e x1 socket 
for this card.



But why are you buying 16 x 2TB disks, if not for performance?


I'm not, I'll only have 6 of them when the rest get here.  One of them 
may go into one of my milling machines to replace the last spinning rust 
on the property.



I've never heard of Gigastone and can offer no assessment of
them.
Relatively new on amazon. A month ago nearly the only 2T available, in a 
month the 2T list has been had the decimal point shifted right at least 
1 place. And the prices have risen 10 bucks. $89/copy today.


-dsr-
.

Take care and stay well, Dan.

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, 1940)
If we desire respect for the law, we must first make the law respectable.
 - Louis D. Brandeis
Genes Web page 

Re: sata driver compataility Q

[David Christensen]

On 9/15/23 12:28, gene heskett wrote:

Greetings all;

I've just ordered some stuff to rebuild or expand my Raid setup.
This 16 port sata-III pci-e card:

along with a bigger drive cage, cables and such and some gigastone 2T 
drives to make a raid big enough to run amanda. And maybe put a new card 
in front of my 2T /home raid10.


The card claims linux compatibility.

Can anyone advise me on the gotcha's of such a 16 port beast?

Thanks all.

Cheers, Gene Heskett.



Searching Amazon for "gigastone 2T", I see:


https://www.amazon.com/Gigastone-Internal-Compatible-Desktop-Laptop/dp/B0BN5978X1

540 MB per second


PCIe 3.0 x1 is rated for 985 MB/s.

https://en.wikipedia.org/wiki/Pcie


So, the PCIe 3.0 x1 connector is going to be a bottleneck when accessing 
more than one SSD.



I suggest that you pick an HBA with a wider PCIe connector -- PCIe 3.0 
x8 (7.88 GB/s) is a reasonable match for sixteen SSD's (8.64 GB/s). 
PCIe 3.0 x16 would eliminate the PCIe bottleneck.



Is everything going into one chassis?  Have you considered an external 
drive chassis?



Make sure your power supply(s) are adequate to the task.


David

sata driver compataility Q

[gene heskett]

Greetings all;

I've just ordered some stuff to rebuild or expand my Raid setup.
This 16 port sata-III pci-e card:

along with a bigger drive cage, cables and such and some gigastone 2T 
drives to make a raid big enough to run amanda. And maybe put a new card 
in front of my 2T /home raid10.


The card claims linux compatibility.

Can anyone advise me on the gotcha's of such a 16 port beast?

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, 1940)
If we desire respect for the law, we must first make the law respectable.
 - Louis D. Brandeis
Genes Web page 

Re: [IMPORTANT] Debian images no longer works for GPU driver installation with apt update

[Andrew M.A. Cater]

On Tue, Aug 29, 2023 at 10:53:35AM -0700, Wenyan Hu wrote:
> Hi Team,
> 
> Google's GPU driver installation no longer works for Debian images since
> 08/23/2023.
> 
> Google is using
> https://cloud.google.com/compute/docs/gpus/install-drivers-gpu#installation_scripts
> to
> install GPU drivers on Debian images. It executes scripts as
> https://raw.githubusercontent.com/GoogleCloudPlatform/compute-gpu-installation/main/linux/install_gpu_driver.py
> .
> 

Maybe take that up with Google - their script and their engineers to fix?

> Now for the latest images, if running scripts as below:
> 
> ```
> $ apt-cache search linux-headers | grep -i $(uname -r)
> $ apt update
> $apt install -y linux-headers-5.10.0-24-cloud-amd64
> software-properties-common pciutils gcc make dkms
> ```
> After running `apt update`, we find the package
> linux-headers-5.10.0-24-cloud-amd64 no longer exists, which results in the
> `apt install linux-headers-5.10.0-24-cloud-amd64` failure.
> 

apt update and then apt upgrade, maybe, to bring the system up to date
rather than just updating the list of available packages?

> It seems after the `apt update`, the index somehow no longer points to
> linux-headers-5.10.0-23-cloud-amd64 and linux-headers-5.10.0-24-cloud-amd64.
> 
> I know it would work if we (1) not do `apt update`, or (2) install newer
> kernel packages.

If you *only* do apt update, nothing has changed.

> But for (1), the apt update is required for our other package
> installations, (2) the kernel package update needs VM rebooting.
> 

This is always likely to be the case: if you do a kernel package update
you *should* reboot to the new kernel, since this is likely to have fixed
security or other problems.

> It is now blocking my team's image release and product functionality
> because we need the GPU driver installation without VM rebooting.
> 

As above, you will need to reboot the VM eventually.

> Could you please prioritize the issue and provide some tips or support?
> 

This is an end user support and information list: we can't provide an
immediate fix to the issue. I would have suggested the debian cloud image
team if I hadn't already suggested taking this up with Google.

> Thanks,
> Wenyan

With every good wish,

Andy Cater

Re: Debian images no longer works for GPU driver installation with apt update

[Michael Kjörling]

On 29 Aug 2023 10:53 -0700, from wen...@google.com (Wenyan Hu):
> Hi Team,

You're in the wrong place. This is a mailing list for Debian _users_
willing to help each other out. Some subscribers might be officially
involved with the Debian project in various capacities, but that's not
the primary purpose of the debian-user mailing list.


> $ apt-cache search linux-headers | grep -i $(uname -r)
> $ apt update
> $apt install -y linux-headers-5.10.0-24-cloud-amd64

It looks like you might be relying on package versions being the same
before and after running apt update. That is an invalid assumption.


> Could you please prioritize the issue and provide some tips or support?

If something within the Debian distribution is broken, see
https://www.debian.org/Bugs/ for how to file a bug report.

However, that your script is apparently expecting a specific version
of a kernel package, or making assumptions about what changes happen
as part of an apt{,-get} update, hardly seems to me as though
something in Debian is broken. That seems more likely to be a problem
with assumptions made in your script. Kernel packages are upgraded
regularly in this manner.

See also http://www.catb.org/esr/faqs/smart-questions.html#urgent

-- 
Michael Kjörling  https://michael.kjorling.se
“Remember when, on the Internet, nobody cared that you were a dog?”

[IMPORTANT] Debian images no longer works for GPU driver installation with apt update

[Wenyan Hu]

Hi Team,

Google's GPU driver installation no longer works for Debian images since
08/23/2023.

Google is using
https://cloud.google.com/compute/docs/gpus/install-drivers-gpu#installation_scripts
to
install GPU drivers on Debian images. It executes scripts as
https://raw.githubusercontent.com/GoogleCloudPlatform/compute-gpu-installation/main/linux/install_gpu_driver.py
.

Now for the latest images, if running scripts as below:

```
$ apt-cache search linux-headers | grep -i $(uname -r)
$ apt update
$apt install -y linux-headers-5.10.0-24-cloud-amd64
software-properties-common pciutils gcc make dkms
```
After running `apt update`, we find the package
linux-headers-5.10.0-24-cloud-amd64 no longer exists, which results in the
`apt install linux-headers-5.10.0-24-cloud-amd64` failure.

It seems after the `apt update`, the index somehow no longer points to
linux-headers-5.10.0-23-cloud-amd64 and linux-headers-5.10.0-24-cloud-amd64.

I know it would work if we (1) not do `apt update`, or (2) install newer
kernel packages.
But for (1), the apt update is required for our other package
installations, (2) the kernel package update needs VM rebooting.

It is now blocking my team's image release and product functionality
because we need the GPU driver installation without VM rebooting.

Could you please prioritize the issue and provide some tips or support?

Thanks,
Wenyan

Re: Bookworm boot stacks with black screen after NVIDIA driver installed.

[Juan R. de Silva]

On Wed, 16 Aug 2023 15:31:52 +0300, Махно wrote:

> Hello. Just stick with open source video driver nouveau.
This is what I'm doing right now. However it performes, as I already said, 
"good enough". Meaning there are some problems. It freezes the system from 
time to time. Well, for now I do not have any choice anyway.

Re: Nvidia 390 driver no longer available for Bookworm; nouveau constantly freezes. Solutions?

[Alexander V. Makartsev]

On 17.08.2023 20:48, Luiz Romário Santana Rios wrote:


Hello, all,

(Please cc me when replying, I'm not subscribed to the mailing list)

I have (according do lspci) a NVIDIA Corporation GK208B [GeForce GT 
710] (rev a1) graphics card. The correct proprietary driver for this 
card seems to be the Nvidia 390 series.


When Bookworm came out, I was running Bullseye with this driver and 
never really had any graphical issues that I can remember. Upon 
upgrading to Bullseye, I eventually found out this driver was no 
longer available and I was stuck with nouveau. I was recommended the 
tesla 470 driver during the upgrade, but I never got it to work.


According to Nvidia's official website GT 710 is supported up to 470 
version. [1]
I've never tried to install "tesla" flavor of nvidia-driver, so I can't 
suggest anything about it.
When I needed a specific version I always build a simple backport [2] 
from "testing" repo, so I'd try to do the same for a 470.199.02 version, 
which is now available from "oldstable-proposed-updates" repo.
You can also try the official package from Nvidia website, at least to 
test if 470 version works and solves your issue with freezing.


The nouveau driver would be fine, since I don't do anything 
graphics-intensive like gaming and I get the bonus of being able to 
run Wayland, which is cool. Except that I started noticing constant 
freezes for no apparent reason. They usually take a few hours to 
happen, sometimes I can spend a day or maybe two without freezes, but 
they _keep happening_. I'm gonna spare you of the details of the 
effort I spent trying to investigate this problem, but what I found 
out is:


  * It happens in any of the kernels I have installed: 6.1, 6.0, 5.11, 4.9
  * It happens in Plasma, and it doesn't matter if I'm running X11 or
Wayland
  * It _seems_ not to happen under GNOME Wayland, but I think it
happens under GNOME X11
  * It appears to be related with screen sharing, because:
  o When I was using Plasma, I had the impression that the
graphics froze way more often when I was sharing my screen
  o Screen sharing doesn't work on GNOME Wayland (maybe related to
why it didn't freeze yet)
  o The one time where I tried to use GNOME X11, it froze
immediately after I tried to share my screen
  * The kernel logs seem to indicate some missing firmware right
before the graphics card freezes (and indeed some firmware is missing)

I had issues very similar to your description. Even made a bug report 
[3] about it.
I don't have any instability or "freezing" issues with recent kernel and 
driver versions, but I've also disabled hardware acceleration for my 
browser as a workaround and haven't actually checked if the issue is gone.
Somehow it was all tied to video hardware acceleration and DE 
Compositor, which I kept disabled. I was able to repro the issue quite 
reliably, back then.
I've thought this workaround was better than downgrading driver version 
to 416 and stick to it forever.
Now I've upgraded system to Bookworm, and use 525.105.17 version, 
delaying update to an up-to-date version from stable, and haven't seen 
"Xid" errors or "freezes" for many months, Compositor is enabled, games 
and movies play without any issues.


Screen sharing is essential for me to work, so this is really not a 
great situation to be in. For comparison, my laptop, which was also 
updated to Bookworm in the same day I updated this PC, had none of the 
problems I'm describing here, since it has an integrated intel 
graphics card.


I appreciate that Nvidia cards can be very problematic and that the 
one card I own is really old, but I didn't have this problem until 
updating and I know this card can work just fine with the right 
driver. Unfortunately, right now, nouveau is not adequate for me to 
work, but the driver that does work is not officially available. What 
should I do?


  * Should I try updating to a newer kernel to see if nouveau got fixed?

I haven't checked, but I think nouveau didn't had a commit in years. It 
is an abandon-ware for me.



  * Should I try installing the missing firmware?


I'd try that, at least to see if it solves the issue.
Keep in mind, non-free firmware was separated to another repo in 
Bookworm. [4]



[1] https://www.nvidia.com/Download/driverResults.aspx/205995/en-us/
[2] https://wiki.debian.org/SimpleBackportCreation
[3] https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1016542
[4] https://wiki.debian.org/SourcesList

--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Nvidia 390 driver no longer available for Bookworm; nouveau constantly freezes. Solutions?

[Luiz Romário Santana Rios]

Hello, all,

(Please cc me when replying, I'm not subscribed to the mailing list)

I have (according do lspci) a NVIDIA Corporation GK208B [GeForce GT 710] 
(rev a1) graphics card. The correct proprietary driver for this card 
seems to be the Nvidia 390 series.


When Bookworm came out, I was running Bullseye with this driver and 
never really had any graphical issues that I can remember. Upon 
upgrading to Bullseye, I eventually found out this driver was no longer 
available and I was stuck with nouveau. I was recommended the tesla 470 
driver during the upgrade, but I never got it to work.


The nouveau driver would be fine, since I don't do anything 
graphics-intensive like gaming and I get the bonus of being able to run 
Wayland, which is cool. Except that I started noticing constant freezes 
for no apparent reason. They usually take a few hours to happen, 
sometimes I can spend a day or maybe two without freezes, but they _keep 
happening_. I'm gonna spare you of the details of the effort I spent 
trying to investigate this problem, but what I found out is:


 * It happens in any of the kernels I have installed: 6.1, 6.0, 5.11, 4.9
 * It happens in Plasma, and it doesn't matter if I'm running X11 or
   Wayland
 * It _seems_ not to happen under GNOME Wayland, but I think it happens
   under GNOME X11
 * It appears to be related with screen sharing, because:
 o When I was using Plasma, I had the impression that the graphics
   froze way more often when I was sharing my screen
 o Screen sharing doesn't work on GNOME Wayland (maybe related to
   why it didn't freeze yet)
 o The one time where I tried to use GNOME X11, it froze
   immediately after I tried to share my screen
 * The kernel logs seem to indicate some missing firmware right before
   the graphics card freezes (and indeed some firmware is missing)

Screen sharing is essential for me to work, so this is really not a 
great situation to be in. For comparison, my laptop, which was also 
updated to Bookworm in the same day I updated this PC, had none of the 
problems I'm describing here, since it has an integrated intel graphics 
card.


I appreciate that Nvidia cards can be very problematic and that the one 
card I own is really old, but I didn't have this problem until updating 
and I know this card can work just fine with the right driver. 
Unfortunately, right now, nouveau is not adequate for me to work, but 
the driver that does work is not officially available. What should I do?


 * Should I try updating to a newer kernel to see if nouveau got fixed?
 * What it I try installing the 390 driver from Bullseye in my Bookworm
   installation? Would that cause any problems?
 * Should I try installing the missing firmware?
 * Any other ideas?

Thank you for the help!

Re: Bookworm boot stacks with black screen after NVIDIA driver installed.

[Махно]

Hello. Just stick with open source video driver nouveau.

2023-08-16, tr, 04:13 Alexander V. Makartsev  rašė:
>
> On 16.08.2023 04:06, Juan R.D. Silva wrote:
>
> Hi folks.
>
> Fresh Bookworm install on Dell M4800 Precision with i7-4810MQ CPU @ 2.80GHz 
> and NVIDIA Quadro K2100M graphic card. No problems install, the system works 
> with the default nouveau driver well enough.
>
> My video card is supported according to NVIDIA.
>
> Information on official Nvidia website suggests otherwise. Last driver 
> version that supports your VGA is 418 release. [1]
> Newer driver versions don't mention K2100M on "Supported Products" list.
>
> Debian nvidia-detect advised installation of nvidia-tesla-470-driver 
> available in repo. After installing it the system boot stacks somewhere in 
> the middle with a weird black screen. No access to TTYs, no cursor, no 
> reaction to keyboard. I tried to boot with "nomodeset" in GRUB with no result.
>
> Installation of drivers newer than 418 version wouldn't work. The only thing 
> they will do is blacklist "nouveau".
> Version 418 is available in Buster which is current "oldoldstable". [2]
> You can try to backport driver from "oldoldstable" to Bookworm. It might be 
> an impossible task due to possible incompatibilities with kernel 6.x
> There is also an option to install official driver package from Nvidia.
> Or you can install and use Buster.
>
>
> [1] https://www.nvidia.com/download/driverResults.aspx/153717/en-us/
> [2] https://wiki.debian.org/DebianOldOldStable
>
> --
> With kindest regards, Alexander.
>
> ⢀⣴⠾⠻⢶⣦⠀
> ⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
> ⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
> ⠈⠳⣄

Re: Bookworm boot stacks with black screen after NVIDIA driver installed.

[Alexander V. Makartsev]

On 16.08.2023 04:06, Juan R.D. Silva wrote:

Hi folks.

Fresh Bookworm install on Dell M4800 Precision with i7-4810MQ CPU @ 
2.80GHz and NVIDIA Quadro K2100M graphic card. No problems install, 
the system works with the default nouveau driver well enough.


My video card is supported according to NVIDIA. 
Information on official Nvidia website suggests otherwise. Last driver 
version that supports your VGA is 418 release. [1]

Newer driver versions don't mention K2100M on "Supported Products" list.

Debian nvidia-detect advised installation of nvidia-tesla-470-driver 
available in repo. After installing it the system boot stacks 
somewhere in the middle with a weird black screen. No access to TTYs, 
no cursor, no reaction to keyboard. I tried to boot with "nomodeset" 
in GRUB with no result.
Installation of drivers newer than 418 version wouldn't work. The only 
thing they will do is blacklist "nouveau".

Version 418 is available in Buster which is current "oldoldstable". [2]
You can try to backport driver from "oldoldstable" to Bookworm. It might 
be an impossible task due to possible incompatibilities with kernel 6.x

There is also an option to install official driver package from Nvidia.
Or you can install and use Buster.


[1] https://www.nvidia.com/download/driverResults.aspx/153717/en-us/
[2] https://wiki.debian.org/DebianOldOldStable

--
With kindest regards, Alexander.

⢀⣴⠾⠻⢶⣦⠀
⣾⠁⢠⠒⠀⣿⡁ Debian - The universal operating system
⢿⡄⠘⠷⠚⠋⠀ https://www.debian.org
⠈⠳⣄

Bookworm boot stacks with black screen after NVIDIA driver installed.

[Juan R.D. Silva]

Hi folks.

Fresh Bookworm install on Dell M4800 Precision with i7-4810MQ CPU @ 
2.80GHz and NVIDIA Quadro K2100M graphic card. No problems install, the 
system works with the default nouveau driver well enough.


My video card is supported according to NVIDIA. Debian nvidia-detect 
advised installation of nvidia-tesla-470-driver available in repo. After 
installing it the system boot stacks somewhere in the middle with a 
weird black screen. No access to TTYs, no cursor, no reaction to 
keyboard. I tried to boot with "nomodeset" in GRUB with no result.


When rebooted in restore mode neither journal -b or journal -xe showed 
any problems. Nvidia module seems to be inserted and loaded. No other 
problems reported. The only way out is to execute 'apt purge "*nvidia*", 
and to reboot. The system then reverts to nouveau driver without any 
problems.


Anyone, please?

Thanks.

Re: Debian as daily driver; WiFi networking and firmware (was: General Questions)

[Michael Kjörling]

On 25 Jul 2023 18:26 +0600, from rifesourcec...@gmail.com (Source Code):
> Using Debian for PC OS is not good? Is it recommended only for servers?

Debian is entirely usable as a daily driver workstation OS. I've been
using it as such for around a decade, possibly longer; I have old
notes and Debian packages dating back to wheezy on my current desktop
system, and I'm quite sure that's not when I started using Debian
specifically.

Debian will, however, _also_ work very well for servers; and many of
the choices one might make to reduce memory footprint (such as not
running a GUI) lend themselves better toward a server installation
than to a workstation setup.


> It turns out you need free firmware to use wifi? But I can use wifi, but
> only with some DE. I just can't use it just from the start without DE.

"Non-free" firmware. Often readily distributable, but does not meet
the Debian Free Software Guidelines (DFSG).

One big change in Debian 12 was to split such firmware out of the
non-free component (where it has been provided since, it appears,
Debian 6/Squeeze) into its own component named non-free-firmware.

Do you have Network Manager installed? Try running "nmcli connection
show" and "iwconfig" when logged in. Do those work?

-- 
Michael Kjörling  https://michael.kjorling.se
“Remember when, on the Internet, nobody cared that you were a dog?”

Re: Driver r3189 pour carte PCI réseau DLink DGE-528T : jouable ?

[didier gaumet]

Le 18/07/2023 à 01:24, roger.tar...@free.fr a écrit :




*De: *"NoSpam" 

[...]
De nos jours le réseau se gère par systemd-networkd ou cloud-init ou 
netplan (Ubuntu) ou ...


Comme l'indique NoSpam certaines méthodes sont privilégiées de nos 
jours: celle que privilégie Debian (systemd-networkd) pour un serveur et 
sa mise en oeuvre sont indiquées dans la manuel de référence Debian:

https://www.debian.org/doc/manuals/debian-reference/ch05.fr.html#_the_modern_network_configuration_without_gui

[...]



[...]
Qui est en charge de la gestion de cette page avec qui on puisse en 
discuter ?

[...]

c'est un wiki donc par définition une plate-forme de rédaction 
collaborative non-hiérarchisée où chacun peut rédiger un article ou le 
modifier.
Par définition aussi, donc, que le contenu d'un wiki soit actualisé 
régulièrement, ou tout simplement pertinent, n'est pas garanti (encore 
moins que dans un projet de documentation supervisé: en gros un projet 
doc va être pertinent ou non suivant les compétences de son responsable, 
alors qu'un wiki peut contenir des articles merveilleux d'intelligence 
côtoyant d'autres articles qui sont des abîmes de bêtise)  .
Donc il peut être judicieux de regarder la documentation d'un projet 
plutôt que son wiki, ou à tout le moins avant, lorsque les deux choses 
sont séparées, ce qui est le cas de Debian

Re: Driver r3189 pour carte PCI réseau DLink DGE-528T : jouable ?

[roger . tarani]

De: "NoSpam"  
À: "Liste Debian"  
Envoyé: Lundi 17 Juillet 2023 22:34:41 
Objet: Re: Driver r3189 pour carte PCI réseau DLink DGE-528T : jouable ? 




Le 17/07/2023 à 19:49, [ mailto:roger.tar...@free.fr | roger.tar...@free.fr ] a 
écrit : 


[...] 
La dernière fois, j'avais galéré avec la documentation debian antique (que je 
prenais pour récente...) et avec les tutos non datés.
En fait, NM était installé (debian 9 ou 10) et toutes les invitations des tutos 
et docs à configurer "facilement" la machine avec 
les resolv.conf, resolvconf, et autres /etc/network/interfaces ne servaient à 
rien. Puisque c'était géré par NM ! 




Faux. 
NM ne gère PAS les interfaces filaires définies dans /etc/network/interfaces. 
Aussi la résolution des noms est gérée par systemd-resolved ou resolvconf qui 
sont indépendants de NM. 
Enfin NM est utilisé majoritairement pour les stations de travail, pour les 
serveurs on s'en passe (perso je le retire et lui préfère ifupdown). 
De nos jours le réseau se gère par systemd-networkd ou cloud-init ou netplan 
(Ubuntu) ou ... 
[ https://wiki.debian.org/fr/NetworkManager | 
https://wiki.debian.org/fr/NetworkManager ] 


Extrait: "...Bien qu'il ait été initialement destiné aux ordinateurs de bureau, 
il a plus récemment été choisi comme logiciel de gestion de réseau par défaut 
dans certaines [ 
https://access.redhat.com/documentation/en-US/Red_Hat_Enterprise_Linux/7/html/Networking_Guide/sec-NetworkManager_and_the_Network_Scripts.html
 | distributions Linux orientées serveur hors Debian. ] " 




Merci pour tes précisions très utiles pour moi qui ait un niveau moyen en 
réseau 
(j'arrive à faire marcher le réseau pour connecter un hôte à un LAN, à 
internet; mais il ne faut pas me demander de faire du routage sophistiqué ou 
alors de faire un comparatif de multiples outils). 




La page [ https://wiki.debian.org/fr/NetworkManager | 
https://wiki.debian.org/fr/NetworkManager ] , et celles liées sont typiquement 
celles que j'avais pu parcourir il y a quelques années, et qui me faisaient me 
poser beaucoup plus de questions que je n'en avais en arrivant. 
D'ailleurs, le lien [ 
https://access.redhat.com/documentation/en-US/Red_Hat_Enterprise_Linux/7/html/Networking_Guide/sec-NetworkManager_and_the_Network_Scripts.html
 | distributions Linux orientées serveur hors Debian ] est en erreur 404, ce 
qui colle un doute sur la fraîcheur de la page. 


A ce propos : à qui signaler ce lien cassé ? 





Je vois que NM peut quand même gérer les interfaces activées dans 
/etc/network/interfaces si on a managed= true dans 
/etc/NetworkManager/NetworkManager.conf 
Pourquoi pas... 

Je lis aussi : 
Redémarrez NetworkManager : 
sudo systemctl restart NetworkManager.service 


À partir de Debian 11 « Bullseye », utilisez : 
sudo service NetworkManager restart 

C'est bizarre (pour moi) car historiquement, j'ai appris à utiliser la commande 
service avant de passer à systemctl. Ce qui colle encore un doute sur la page. 


Si on cherche la dernière date de modification, on peut trouver un tout petit 
"( [ https://wiki.debian.org/fr/NetworkManager?action=info | last modified 
2021-04-28 ] )" en pied de page. Sa place serait en entête. 

Qui est en charge de la gestion de cette page avec qui on puisse en discuter ? 


Cette page pointe vers : [ https://wiki.debian.org/fr/NetworkConfiguration | 
https://wiki.debian.org/fr/NetworkConfiguration ] 
qui est assez bien faite mais il y en a une trop longue tartine avec pleins de 
références, notamment à resolvconf (à ne pas confondre avec le fichier 
resolv.conf ...) 


Je lis : Comment créer une connexion tolérante aux pannes avec un vlan ( dans 
Etch jusqu'à Stretch ) (debian 4 à 9) 

En tant que débutant j'étais perdu avec ces pages. 
Aujourd'hui, je suis assez mal à l'aise en lisant cette page qui parle de trucs 
dépassés ("de debian 4/Etch à debian 9/Stretch") et où il semble qu'il y ait du 
désordre. 

Je crois qu'il faudrait deux niveaux de lecture : 
- un niveau pour debian 11 ou 12 (une recette de base courte avec quelques 
variations; - pour un serveur ; - pour une machine de bureau. Un peu dans 
l'esprit des pages d'Arch Linux, très carrées) 
- un niveau "historique/vieilleries" (tout ce qui concerne d'anciennes versions 
et qui explique l'évolution) 

A nouveau : qui est en charge de la gestion de cette page avec qui on puisse en 
discuter ? 

Mes observations sont un peu longues mais ça peut peut-être servir à améliorer 
la documentation debian. 

== 
En résumé : 
Une fois que j'ai installé les drivers pour les cartes réseau installées, 
- sur un serveur, je désactive NM, je définis mes interfaces (static ou dhcp) 
dans /etc/network/interfaces et j'utilise ifupdown (plus le routage entre 
réseaux, si nécessaire) 
- sur une machine de bureau, je peux utiliser NM ou systemd-networkd 
(https://wiki.debian.org/SystemdNetworkd et 
https://wiki.archlinux.org/title/Syste

Re: Driver r3189 pour carte PCI réseau DLink DGE-528T : jouable ?

[NoSpam]

Le 17/07/2023 à 19:49, roger.tar...@free.fr a écrit :

[...]
La dernière fois, j'avais galéré avec la documentation debian antique (que je 
prenais pour récente...) et avec les tutos non datés.
En fait, NM était installé (debian 9 ou 10) et toutes les invitations des tutos et docs à 
configurer "facilement" la machine avec les resolv.conf, resolvconf, et autres 
/etc/network/interfaces ne servaient à rien. Puisque c'était géré par NM !


Faux. NM ne gère PAS les interfaces filaires définies dans 
/etc/network/interfaces. Aussi la résolution des noms est gérée par 
systemd-resolved ou resolvconf qui sont indépendants de NM Enfin NM est 
utilisé majoritairement pour les stations de travail, pour les serveurs 
on s'en passe (perso je le retire et lui préfère ifupdown). De nos jours 
le réseau se gère par systemd-networkd ou cloud-init ou netplan (Ubuntu) 
ou ...


https://wiki.debian.org/fr/NetworkManager

Extrait: "...Bien qu'il ait été initialement destiné aux ordinateurs de 
bureau, il a plus récemment été choisi comme logiciel de gestion de 
réseau par défaut dans certaines distributions Linux orientées serveur 
hors Debian. 
"

Re: Driver r3189 pour carte PCI réseau DLink DGE-528T : jouable ?

[roger . tarani]

- Mail original -
De: "Jean-François Bachelet" 
À: "Liste Debian" 
Envoyé: Lundi 17 Juillet 2023 18:03:37
Objet: Re: Driver r3189 pour carte PCI réseau DLink DGE-528T : jouable ?

hello :)


Le 17/07/2023 à 15:45, roger.tar...@free.fr a écrit :
>
>
> Le 17 juil. 2023 à 15:01, Bernard Schoenacker
>  a écrit :
>
> Bonjour Roger,
>
> Pourquoi vouloir rédiger un tel billet qui contient qu'un ensemble
> de données bonnes à jeter à la poubelle ?
>
> Ce qui est viable :
> lspci -vv |grep -i ethernet
> module recherché : realtek 8169
> recherche sur la machine :
> find /lib/modules -name r8169.ko
> réponse :
> /lib/modules/6.3.0-1-amd64/kernel/drivers/net/ethernet/realtek/r8169.ko
>
> Les dénominations commerciales ne sont là que comme dans la
> réclame où c'est la marmotte qui emballe le chocolat dans
> du papier aluminium...
>
> Que contient le fichier sources.list ?
> Intégration des mots : non-free-firmware
>
> deb [arch=amd64]  https://deb.debian.org/debian bookworm main
> contrib non-free non-free-firmware
> deb [arch=amd64] https://deb.debian.org/debian/
> bookworm-proposed-updates contrib main non-free non-free-firmware
> deb [arch=amd64] https://deb.debian.org/debian bookworm-updates
> main contrib non-free non-free-firmware
> deb [arch=amd64] https://deb.debian.org/debian-security/
> bookworm-security main contrib non-free non-free-firmware
>
> Je n'apprécie guère l'esbroufe sauce Microsoft admin...
>
> « Ce que l’on conçoit bien s’énonce clairement »
> Nicolas Boileau
>
>
> Merci pour l'aimable question posée sur cette liste et continuez à
> vous instruire
> convenablement
>
> Bien à vous
>
> Bernard
>
> Bonjour Nicolas,
> Merci pour ton aide.
> Je ne sais pas si tu réponds à mon message initial ou au PS avec 
> quelques commandes qui semblaient pourtant prometteuses, pour un amateur.
> Je t'ai écouté. J'ai donc tout jeté à la poubelle, à commencer par 
> cette vieille carte réseau sans grand intérêt pour personne !

heu ! c'est quand même une carte Gigabit ethernet la 8169... donc loin 
d'être inutile ;)

jeff

Merci.
J'ai donc vidé les poubelles.

$ lspci -vv
10:00.0 Ethernet controller: D-Link System Inc DGE-528T Gigabit Ethernet 
Adapter (rev 10)
Subsystem: D-Link System Inc DGE-528T PCI Gigabit Ethernet Adapter
Physical Slot: 1-2
Control: I/O+ Mem+ BusMaster- SpecCycle- MemWINV- VGASnoop- ParErr- 
Stepping- SERR+ FastB2B- DisINTx-
Status: Cap+ 66MHz+ UDF- FastB2B+ ParErr- DEVSEL=medium >TAbort- 
SERR- 
Kernel modules: r8169


$ sudo nmcli device status 
DEVICE   TYPE  STATE   CONNECTION   
br0  bridgeconnected   br0  
docker0  bridgeconnected (externally)  docker0  
virbr0   bridgeconnected (externally)  virbr0   
enp0s25  ethernet  connected   bridge-slave-enp0s25 
vnet0tun   connected (externally)  vnet0
lo   loopback  unmanaged   --  


Je crois lire que la carte est reconnue.
Et il doit suffit de configurer l'interface.

Encore une bonne révision, puisque je fais ça rarement...

La dernière fois, j'avais galéré avec la documentation debian antique (que je 
prenais pour récente...) et avec les tutos non datés.
En fait, NM était installé (debian 9 ou 10) et toutes les invitations des tutos 
et docs à configurer "facilement" la machine avec les resolv.conf, resolvconf, 
et autres /etc/network/interfaces ne servaient à rien. Puisque c'était géré par 
NM !
J'en avais perdu des heures à tourner en rond et à m'arracher les cheveux.
Tous ces articles auraient du être brûlés sur un bûcher ! Et moi avec pour y 
avoir cru. 


Du coup, je vous pose la question : 
Quelle est la manière directe de configurer cette carte réseau (avec nmcli) ?

Par exemple, est-ce que ce  tuto est valable ? 
https://sysreseau.net/networkmanager-nmcli-nmtui/


Pourquoi j'ajoute une deuxième carte réseau ?
J'aimerais expérimenter du routage et du fw + vpn server + squid pour mettre à 
jour mes connaissances en réseau. 

Merci.

Re: Driver r3189 pour carte PCI réseau DLink DGE-528T : jouable ?

[Jean-François Bachelet]

hello :)


Le 17/07/2023 à 15:45, roger.tar...@free.fr a écrit :



Le 17 juil. 2023 à 15:01, Bernard Schoenacker
 a écrit :

Bonjour Roger,

Pourquoi vouloir rédiger un tel billet qui contient qu'un ensemble
de données bonnes à jeter à la poubelle ?

Ce qui est viable :
lspci -vv |grep -i ethernet
module recherché : realtek 8169
recherche sur la machine :
find /lib/modules -name r8169.ko
réponse :
/lib/modules/6.3.0-1-amd64/kernel/drivers/net/ethernet/realtek/r8169.ko

Les dénominations commerciales ne sont là que comme dans la
réclame où c'est la marmotte qui emballe le chocolat dans
du papier aluminium...

Que contient le fichier sources.list ?
Intégration des mots : non-free-firmware

deb [arch=amd64]  https://deb.debian.org/debian bookworm main
contrib non-free non-free-firmware
deb [arch=amd64] https://deb.debian.org/debian/
bookworm-proposed-updates contrib main non-free non-free-firmware
deb [arch=amd64] https://deb.debian.org/debian bookworm-updates
main contrib non-free non-free-firmware
deb [arch=amd64] https://deb.debian.org/debian-security/
bookworm-security main contrib non-free non-free-firmware

Je n'apprécie guère l'esbroufe sauce Microsoft admin...

« Ce que l’on conçoit bien s’énonce clairement »
Nicolas Boileau


Merci pour l'aimable question posée sur cette liste et continuez à
vous instruire
convenablement

Bien à vous

Bernard

Bonjour Nicolas,
Merci pour ton aide.
Je ne sais pas si tu réponds à mon message initial ou au PS avec 
quelques commandes qui semblaient pourtant prometteuses, pour un amateur.
Je t'ai écouté. J'ai donc tout jeté à la poubelle, à commencer par 
cette vieille carte réseau sans grand intérêt pour personne !


heu ! c'est quand même une carte Gigabit ethernet la 8169... donc loin 
d'être inutile ;)


jeff

Re: Driver r3189 pour carte PCI réseau DLink DGE-528T : jouable ?

[roger . tarani]

Le 17 juil. 2023 à 15:01, Bernard Schoenacker  a 
écrit : 





BQ_BEGIN

Bonjour Roger, 

Pourquoi vouloir rédiger un tel billet qui contient qu'un ensemble 
de données bonnes à jeter à la poubelle ? 

Ce qui est viable : 
lspci -vv |grep -i ethernet 
module recherché : realtek 8169 
recherche sur la machine : 
find /lib/modules -name r8169.ko 
réponse : 
/lib/modules/6.3.0-1-amd64/kernel/drivers/net/ethernet/realtek/r8169.ko 

Les dénominations commerciales ne sont là que comme dans la 
réclame où c'est la marmotte qui emballe le chocolat dans 
du papier aluminium... 

Que contient le fichier sources.list ? 
Intégration des mots : non-free-firmware 

deb [arch=amd64] https://deb.debian.org/debian bookworm main contrib non-free 
non-free-firmware 
deb [arch=amd64] https://deb.debian.org/debian/ bookworm-proposed-updates 
contrib main non-free non-free-firmware 
deb [arch=amd64] https://deb.debian.org/debian bookworm-updates main contrib 
non-free non-free-firmware 
deb [arch=amd64] https://deb.debian.org/debian-security/ bookworm-security main 
contrib non-free non-free-firmware 

Je n'apprécie guère l'esbroufe sauce Microsoft admin... 

« Ce que l’on conçoit bien s’énonce clairement » 
Nicolas Boileau 


Merci pour l'aimable question posée sur cette liste et continuez à vous 
instruire 
convenablement 

Bien à vous 

Bernard 


BQ_END

Bonjour Nicolas, 
Merci pour ton aide. 
Je ne sais pas si tu réponds à mon message initial ou au PS avec quelques 
commandes qui semblaient pourtant prometteuses, pour un amateur. 
Je t'ai écouté. J'ai donc tout jeté à la poubelle, à commencer par cette 
vieille carte réseau sans grand intérêt pour personne ! 
Merci. 

Re: Driver r3189 pour carte PCI réseau DLink DGE-528T : jouable ?

[Bernard Schoenacker]

Bonjour Roger,

Pourquoi vouloir rédiger un tel billet qui contient qu'un ensemble
de données bonnes à jeter à la poubelle ?

Ce qui est viable : 

lspci -vv |grep -i ethernet

module recherché : realtek 8169

recherche sur la machine :

find /lib/modules -name r8169.ko

réponse :

/lib/modules/6.3.0-1-amd64/kernel/drivers/net/ethernet/realtek/r8169.ko



Les dénominations commerciales ne sont là que comme dans la 
réclame où c'est la marmotte qui emballe le chocolat dans 
du papier aluminium...

Que contient le fichier sources.list ?

Intégration des mots : non-free-firmware


deb [arch=amd64]  https://deb.debian.org/debian bookworm main contrib non-free 
non-free-firmware

deb [arch=amd64] https://deb.debian.org/debian/ bookworm-proposed-updates 
contrib main non-free non-free-firmware

deb [arch=amd64] https://deb.debian.org/debian bookworm-updates main contrib 
non-free non-free-firmware

deb [arch=amd64] https://deb.debian.org/debian-security/ bookworm-security main 
contrib non-free non-free-firmware


Je n'apprécie guère l'esbroufe sauce Microsoft admin...

« Ce que l’on conçoit bien s’énonce clairement »

Nicolas Boileau


Merci pour l'aimable question posée sur cette liste et continuez à vous 
instruire 
convenablement

Bien à vous 

Bernard

Re: Driver r3189 pour carte PCI réseau DLink DGE-528T : jouable ?

[didier gaumet]

A l'instant je suis à la bourre et pas sous Debian, mais regarde dans le 
répertoire /etc/modprobe.d/ si il n'y a pas un fichier qui interdit le 
chargement du module r8169 et vérifie que tu as bien les bon firmwares 
installés (journalctl, chercher les messages de firmware manquants)

Re: Driver r3189 pour carte PCI réseau DLink DGE-528T : jouable ?

[Jean-François Bachelet]

Hello :)

Mets déjà cette carte dans le mac, il devrait la reconnaitre et 
installer le driver pour elle  tout seul ou presque, pourquoi vouloir le 
compiler/installer à la main ?



Jeff

Le 17/07/2023 à 13:50, RogerT a écrit :

Bonjour,

Je voudrais installer une carte réseau additionnelle.
J’ai retrouvé une DLink DGE-528T 
<https://askubuntu.com/questions/39172/how-do-i-install-the-driver-for-a-dlink-dge-528t> achetée 
il y a 5-6 ans, vendue pour mac/linux/win.


Je voudrais savoir si ça vaut le coup de l’installer.

Je vois dans la doc d’origine  « pour linux 2.4 et 2.6 »
Et dans la doc plus récente : aussi noyau 3.16
https://eu.dlink.com/fr/fr/-/media/business_products/dge/dge-528t/datasheet/dge_528t_c1_datasheet_en.pdf

La machine est en 5.10
Je lis ici :
https://linux-hardware.org/index.php?id=pci:1186-4300-1186-4300
Ver Source Config By ID

5.3 - 6.3 drivers/net/ethernet/realtek/r8169_main.c 
<https://github.com/torvalds/linux/blob/master/drivers/net/ethernet/realtek/r8169_main.c> 
CONFIG_ETHERNET CONFIG_NET_VENDOR_REALTEK CONFIG_R8169 1186:4300



J’ai commencé à jouer du lspci et lsmod.
Je ne crois pas voir le driver realtek8169 installé.

J’hésite à me lancer dans une compilation de pilote, etc. vu mes 
expériences parfois difficiles et vu le prix d’une carte réseau récente.


D’après votre expérience, est-ce que ça vaut le coup d’essayer de 
l’installer ?

Et comment ?
Merci.

Re: Driver r3189 pour carte PCI réseau DLink DGE-528T : jouable ?

[roger . tarani]

PS : c'est une occasion de réviser les notions de kernel, de modules et les 
commandes... 

$ find /lib/modules/ -name "*8169*" 
/lib/modules/5.10.0-22-amd64/kernel/drivers/net/ethernet/realtek/r8169.ko 
/lib/modules/5.10.0-23-amd64/kernel/drivers/net/ethernet/realtek/r8169.ko 

$ modprobe r8169 # reconnu, non exécuté) 

$ lsmod | grep real 
snd_hda_codec_realtek 167936 1 
snd_hda_codec_generic 98304 1 snd_hda_codec_realtek 
snd_hda_codec 176128 3 
snd_hda_codec_generic,snd_hda_intel,snd_hda_codec_realtek 
snd_hda_core 110592 4 
snd_hda_codec_generic,snd_hda_intel,snd_hda_codec,snd_hda_codec_realtek 
snd 110592 18 
snd_hda_codec_generic,snd_hwdep,snd_hda_intel,snd_hda_codec,snd_hda_codec_realtek,snd_timer,snd_compress,snd_soc_core,snd_pcm
 


Comment vérifier simplement que le pilote requis (ici : r8169) est bien 
installé? chargé ? 

Je me suis lancé... 


$ lsmod | grep r8169 
(rien) 

$ sudo modprobe r8169 
(rien) 

$ lsmod | grep real 
realtek 24576 0 # < nouveau 
libphy 155648 3 r8169,mdio_devres,realtek 
snd_hda_codec_realtek 167936 1 
snd_hda_codec_generic 98304 1 snd_hda_codec_realtek 
snd_hda_codec 176128 3 
snd_hda_codec_generic,snd_hda_intel,snd_hda_codec_realtek 
snd_hda_core 110592 4 
snd_hda_codec_generic,snd_hda_intel,snd_hda_codec,snd_hda_codec_realtek 
snd 110592 18 
snd_hda_codec_generic,snd_hwdep,snd_hda_intel,snd_hda_codec,snd_hda_codec_realtek,snd_timer,snd_compress,snd_soc_core,snd_pcm
 

$ lsmod | grep r8169 
r8169 102400 0 
mdio_devres 16384 1 r8169 
libphy 155648 3 r8169,mdio_devres,realtek 


De ma petite expérience, j'ai l'impression d'avoir chargé le driver r8169 (pour 
le noyau 5.10.0-23-amd64). 
Ou ai-je oublié quelque chose ? 



De: "roger tarani"  
À: "Liste Debian"  
Envoyé: Lundi 17 Juillet 2023 13:50:45 
Objet: Driver r3189 pour carte PCI réseau DLink DGE-528T : jouable ? 

Bonjour, 

Je voudrais installer une carte réseau additionnelle. 
J’ai retrouvé une [ 
https://askubuntu.com/questions/39172/how-do-i-install-the-driver-for-a-dlink-dge-528t
 | DLink DGE-528T ] achetée il y a 5-6 ans, vendue pour mac/linux/win. 

Je voudrais savoir si ça vaut le coup de l’installer. 

Je vois dans la doc d’origine « pour linux 2.4 et 2.6 » 
Et dans la doc plus récente : aussi noyau 3.16 
[ 
https://eu.dlink.com/fr/fr/-/media/business_products/dge/dge-528t/datasheet/dge_528t_c1_datasheet_en.pdf
 | 
https://eu.dlink.com/fr/fr/-/media/business_products/dge/dge-528t/datasheet/dge_528t_c1_datasheet_en.pdf
 ] 

La machine est en 5.10 
Je lis ici : 
[ https://linux-hardware.org/index.php?id=pci:1186-4300-1186-4300 | 
https://linux-hardware.org/index.php?id=pci:1186-4300-1186-4300 ] 
Ver Source Config By ID 
5.3 - 6.3   drivers/net/ethernet/realtek/ [ 
https://github.com/torvalds/linux/blob/master/drivers/net/ethernet/realtek/r8169_main.c
 | r8169_main.c ] 
CONFIG_ETHERNET CONFIG_NET_VENDOR_REALTEK CONFIG_R8169  1186:4300 

J’ai commencé à jouer du lspci et lsmod. 
Je ne crois pas voir le driver realtek8169 installé. 

J’hésite à me lancer dans une compilation de pilote, etc. vu mes expériences 
parfois difficiles et vu le prix d’une carte réseau récente. 

D’après votre expérience, est-ce que ça vaut le coup d’essayer de l’installer ? 
Et comment ? 
Merci. 

Driver r3189 pour carte PCI réseau DLink DGE-528T : jouable ?

[RogerT]

Bonjour,

Je voudrais installer une carte réseau additionnelle. 
J’ai retrouvé une DLink DGE-528T achetée il y a 5-6 ans, vendue pour 
mac/linux/win. 

Je voudrais savoir si ça vaut le coup de l’installer. 

Je vois dans la doc d’origine  « pour linux 2.4 et 2.6 »
Et dans la doc plus récente : aussi noyau 3.16 
https://eu.dlink.com/fr/fr/-/media/business_products/dge/dge-528t/datasheet/dge_528t_c1_datasheet_en.pdf

La machine est en 5.10
Je lis ici :
https://linux-hardware.org/index.php?id=pci:1186-4300-1186-4300
Ver Source  Config  By ID   
5.3 - 6.3   drivers/net/ethernet/realtek/r8169_main.c   CONFIG_ETHERNET 
CONFIG_NET_VENDOR_REALTEK CONFIG_R8169  1186:4300

J’ai commencé à jouer du lspci et lsmod. 
Je ne crois pas voir le driver realtek8169 installé. 

J’hésite à me lancer dans une compilation de pilote, etc. vu mes expériences 
parfois difficiles et vu le prix d’une carte réseau récente. 

D’après votre expérience, est-ce que ça vaut le coup d’essayer de l’installer ?
Et comment ?
Merci. 

Re: Nvidia graphics driver naming convention

[Anssi Saari]

Nicolas Marie  writes:

> My question is about the package naming: is the 
> nvidia-graphics-drivers-tesla-470 specifically built for Tesla
> graphics, and the fact that I can use it for my consumer Nvidia GPU is
> a positive "side effect"?

The description of package nvidia-tesla-470-driver seems clear on what
GPU families are supported and gives a reference to a documentation file
with more details.

I have no idea about the why of the tesla naming. I'd guess it's a
historical reason and lack of manpower to update.

Re: Nvidia graphics driver naming convention

[Andrew M.A. Cater]

On Sat, Jul 15, 2023 at 09:48:34AM +0200, Nicolas Marie wrote:
> Hello.
> 
> I've got three workstations using different Nvidia GPU. I upgraded them from
> Debian 11.6 to Debian 12:
> 
>  * The first one is a Geforce GT520 (Fermi) so I used the package
>*nvidia-graphics-drivers-legacy-390xx* and it works like a charm.
>  * The second one is a Geforce 1050 Ti (Pascal) si I used
>*nvidia-graphics-drivers* which is the version 525 and it also works
>like a charm.
>  * The last one is a Geforce GT730 (Kepler) so I have to use the Nvidia
>driver version *470*.
> 

https://wiki.debian.org/NvidiaGraphicsDrivers gives the canonical instructions
- I think that one's old enough that nouveau also works well with it too.

> There isn't any package named *nvidia-graphics-drivers-legacy-470xx*, like
> we can find for the 390 mentioned version or the older 340 version. As a
> workaround, I used *nvidia-graphics-drivers-tesla-470* and all the GLX stack
> / SMI utility / nvidia-settings utility. Basically, all seems to be working:
> 
>  * The nvidia module is built through the DKMS workflow and loaded.
>  * The GPU is recognized.
>  * The basic graphics acceleration (KDE windows effects / GLXGears)
>works without any glitch.
> 

See above.

> P.S.: I know that this question type should be escalated during beta states
> but, since I used the Debian Bullseye Backports packages for nearly one
> year, I wrongfully believed these drivers were be maintained in the next
> Debian release.
> 
> Thanks for your answer.
> Regards.*
> *

My standard advice also is to build Nvidia drivers from a bare CLI text only 
install *before* going back to install a desktop environment and preferably 
using
the expert text mode install, but that may well be overkill.

It does, however, prevent nouveau or any other conflicting packages causing
problems if you are determined to use hte proprietary drivers.

As ever, your mileage may vary.

With every good wish,

Andy Cater

Nvidia graphics driver naming convention

[Nicolas Marie]

Hello.

I've got three workstations using different Nvidia GPU. I upgraded them 
from Debian 11.6 to Debian 12:


 * The first one is a Geforce GT520 (Fermi) so I used the package
   *nvidia-graphics-drivers-legacy-390xx* and it works like a charm.
 * The second one is a Geforce 1050 Ti (Pascal) si I used
   *nvidia-graphics-drivers* which is the version 525 and it also works
   like a charm.
 * The last one is a Geforce GT730 (Kepler) so I have to use the Nvidia
   driver version *470*.

There isn't any package named *nvidia-graphics-drivers-legacy-470xx*, 
like we can find for the 390 mentioned version or the older 340 version. 
As a workaround, I used *nvidia-graphics-drivers-tesla-470* and all the 
GLX stack / SMI utility / nvidia-settings utility. Basically, all seems 
to be working:


 * The nvidia module is built through the DKMS workflow and loaded.
 * The GPU is recognized.
 * The basic graphics acceleration (KDE windows effects / GLXGears)
   works without any glitch.

My question is about the package naming: is the 
nvidia-graphics-drivers-tesla-470 specifically built for Tesla graphics, 
and the fact that I can use it for my consumer Nvidia GPU is a positive 
"side effect"? Or is this package totally usable with a consumer GPU 
and, in this case the name *nvidia-graphics-drivers-legacy-470xx* would 
be more relevant for the non-aware debian new users. As a comparison, 
Ubuntu uses a similar naming convention, namely nvidia-driver-390, 
nvidia-driver-470 and nvidia-driver-525.


P.S.: I know that this question type should be escalated during beta 
states but, since I used the Debian Bullseye Backports packages for 
nearly one year, I wrongfully believed these drivers were be maintained 
in the next Debian release.


Thanks for your answer.
Regards.*
*

Re: Fwd: Wayland and NVidia driver conflict

[Marvin Renich]

debian-user and debian-desktop are both good lists for this question.
It is off-topic for debian-devel.

Anyone who answers, please remove debian-devel from the replies.

(Sorry, I don't have an answer for you.)

Thanks...Marvin

Fwd: Wayland and NVidia driver conflict

[Luna Jernberg]

-- Forwarded message -
Från: inkrm 
Date: tors 13 juli 2023 kl 23:48
Subject: Wayland and NVidia driver conflict
To: 


Hi
I just installed Debian 12 'Bookworm' with KDE Plasma 5.27.5 , and I
installed NVidia driver on my machine, following the official guide
(https://wiki.debian.org/NvidiaGraphicsDrivers#Debian_12_.22Bookworm.22).
When I rebooted my machine, I logged in Wayland, but it threw me back
into login manager.
I honestly prefer Wayland because it has better compatibility with 4K
HiDPI displays (which I have), but I can't use it properly, and
instead I have to use X11.
I searched on web but didn't find anything, so I am writing this
email, to ask what do I do.
I'm gonna be very pleased to read your response.

Kind regards, inkrm
--
Sent with Tutanota

Re: Bluetooth driver error on boot after upgrade to bookworm

[The Wanderer]

On 2023-06-16 at 11:00, Unni wrote:

> Hello,
> 
> I am getting Bluetooth firmware error after upgrade. I've tried 
> reinstalling firmware-iwlwifi. But no luck. I've added non-free-firmware 
> in apt source, so thats not the issue. Help me to fix the error on booth.

From that last, I'm presuming that you've rebooted since the upgrade,
and that it's at boot time that the error appears.

> ~# dmesg | egrep -i 'bluetooth'
> [6.051557] Bluetooth: Core ver 2.22
> [6.051579] NET: Registered PF_BLUETOOTH protocol family
> [6.051580] Bluetooth: HCI device and connection manager initialized
> [6.051587] Bluetooth: HCI socket layer initialized
> [6.051589] Bluetooth: L2CAP socket layer initialized
> [6.051593] Bluetooth: SCO socket layer initialized
> [6.093212] bluetooth hci0: firmware: failed to load 
> intel/ibt-20-0-0.sfi (-2)
> [6.093358] bluetooth hci0: firmware: failed to load 
> intel/ibt-20-0-0.sfi (-2)

I see some things online that seem to suggest that this may be a
misleading message, and that the stack may already be using the
available firmware, but is also attempting to load this lower version
and is reporting that that attempt failed. However, those mentions are
somewhat older (2021-ish?), and the messages aren't quite the same, and
it looks like they should have been fixed by now, so I'm not assuming
that they're talking about this same thing.

Can you confirm whether you're seeing any actual functionality issues
along with this change, or whether it's just the fact that the message
is appearing that has you concerned?

FWIW, I get all of the above messages except the two about firmware (I
get two about a command timeout and a "Reading Intel version command
failed" instead)...

> [7.622442] Bluetooth: BNEP (Ethernet Emulation) ver 1.3
> [7.622446] Bluetooth: BNEP filters: protocol multicast
> [7.622450] Bluetooth: BNEP socket layer initialized

...and also get these three...

> [7.623488] Bluetooth: MGMT ver 1.22
> [9.011767] Bluetooth: RFCOMM TTY layer initialized
> [9.011779] Bluetooth: RFCOMM socket layer initialized
> [9.011784] Bluetooth: RFCOMM ver 1.11
> [   32.748460] Bluetooth: hci0: Opcode 0x 401 failed: -16

...but not these. I don't actually know whether Bluetooth functionality
is available on my system, because I don't think I have any hardware
suitable to pair to it. Presumably, however, the lines we have in common
should not be relevant to what you're seeing.


$ apt-file search ibt-20-0-
firmware-iwlwifi: /lib/firmware/intel/ibt-20-0-3.ddc
firmware-iwlwifi: /lib/firmware/intel/ibt-20-0-3.sfi

(This is with the same package version you referenced; it's in both
stable and testing, at the moment.)

So ibt-20-0-3.sfi exists, but not the 20-0-0 version. The question would
then seem to be why the Bluetooth stack is attempting to load that
lower-numbered version.

Given that we're talking about firmware, it may be necessary to
reinitialize the relevant modules (et possibly cetera) after the
upgrade; that can almost certainly be done by rebooting, but should also
be possible to do manually in many cases by removing and reinserting
modules. If (as suggested above) you've rebooted, then presumably this
approach has already been tried.

(In my case, I'm fairly sure I haven't rebooted since the release, but I
was also tracking testing right up to the release so I'm already running
the same kernel etc. that you mentioned.)

-- 
   The Wanderer

The reasonable man adapts himself to the world; the unreasonable one
persists in trying to adapt the world to himself. Therefore all
progress depends on the unreasonable man. -- George Bernard Shaw



signature.asc
Description: OpenPGP digital signature

Bluetooth driver error on boot after upgrade to bookworm

[Unni]

Hello,

I am getting Bluetooth firmware error after upgrade. I've tried 
reinstalling firmware-iwlwifi. But no luck. I've added non-free-firmware 
in apt source, so thats not the issue. Help me to fix the error on booth.



~# dmesg | egrep -i 'bluetooth'
[    6.051557] Bluetooth: Core ver 2.22
[    6.051579] NET: Registered PF_BLUETOOTH protocol family
[    6.051580] Bluetooth: HCI device and connection manager initialized
[    6.051587] Bluetooth: HCI socket layer initialized
[    6.051589] Bluetooth: L2CAP socket layer initialized
[    6.051593] Bluetooth: SCO socket layer initialized
[    6.093212] bluetooth hci0: firmware: failed to load 
intel/ibt-20-0-0.sfi (-2)
[    6.093358] bluetooth hci0: firmware: failed to load 
intel/ibt-20-0-0.sfi (-2)

[    7.622442] Bluetooth: BNEP (Ethernet Emulation) ver 1.3
[    7.622446] Bluetooth: BNEP filters: protocol multicast
[    7.622450] Bluetooth: BNEP socket layer initialized
[    7.623488] Bluetooth: MGMT ver 1.22
[    9.011767] Bluetooth: RFCOMM TTY layer initialized
[    9.011779] Bluetooth: RFCOMM socket layer initialized
[    9.011784] Bluetooth: RFCOMM ver 1.11
[   32.748460] Bluetooth: hci0: Opcode 0x 401 failed: -16

~# hciconfig -a hci0
hci0:    Type: Primary  Bus: USB
    BD Address: 14:F6:D8:D7:FD:C9  ACL MTU: 1021:4  SCO MTU: 96:6
    UP RUNNING PSCAN ISCAN
    RX bytes:1324 acl:0 sco:0 events:122 errors:0
    TX bytes:7670 acl:0 sco:0 commands:116 errors:0
    Features: 0xbf 0xfe 0x0f 0xfe 0xdb 0xff 0x7b 0x87
    Packet type: DM1 DM3 DM5 DH1 DH3 DH5 HV1 HV2 HV3
    Link policy: RSWITCH SNIFF
    Link mode: PERIPHERAL ACCEPT
    Name: 'unni-debian'
    Class: 0x7c0104
    Service Classes: Rendering, Capturing, Object Transfer, Audio, 
Telephony

    Device Class: Computer, Desktop workstation
    HCI Version: 5.2 (0xb)  Revision: 0x2279
    LMP Version: 5.2 (0xb)  Subversion: 0x2279
    Manufacturer: Intel Corp. (2)

~# lsb_release -a
No LSB modules are available.
Distributor ID:    Debian
Description:    Debian GNU/Linux 12 (bookworm)
Release:    12
Codename:    bookworm

~# cat /etc/apt/sources.list
# See https://wiki.debian.org/SourcesList for more information.
deb http://deb.debian.org/debian bookworm main contrib non-free 
non-free-firmware
deb-src http://deb.debian.org/debian bookworm main contrib non-free 
non-free-firmware


deb http://deb.debian.org/debian bookworm-updates main contrib non-free 
non-free-firmware
deb-src http://deb.debian.org/debian bookworm-updates main contrib 
non-free non-free-firmware


deb http://security.debian.org/debian-security/ bookworm-security main 
contrib non-free non-free-firmware
deb-src http://security.debian.org/debian-security/ bookworm-security 
main contrib non-free non-free-firmware

#Testing repo
# deb http://deb.debian.org/debian testing non-free non-free-firmware 
contrib main


~# dpkg-query -W | grep -E '(linux-image|firmware-iwlwifi)'
firmware-iwlwifi    20230210-5
linux-image-5.10.0-10-amd64    5.10.84-1
linux-image-5.10.0-11-amd64    5.10.92-2
linux-image-5.10.0-13-amd64    5.10.106-1
linux-image-5.10.0-14-amd64    5.10.113-1
linux-image-5.10.0-15-amd64    5.10.120-1
linux-image-5.10.0-16-amd64    5.10.127-2
linux-image-5.10.0-17-amd64    5.10.136-1
linux-image-5.10.0-19-amd64    5.10.149-2
linux-image-5.10.0-20-amd64    5.10.158-2
linux-image-5.10.0-21-amd64    5.10.162-1
linux-image-5.10.0-23-amd64    5.10.179-1
linux-image-5.10.0-9-amd64    5.10.70-1
linux-image-6.1.0-9-amd64    6.1.27-1
linux-image-amd64    6.1.27-1

~# uname -mrs
Linux 6.1.0-9-amd64 x86_64

Re: TP Link TL-WN722N adapter & ath9k-htc driver.

[Charles Curley]

On 15 Jun 2023 07:43:01 -0700
pe...@easthope.ca wrote:

> An old Sharp Mebius here has 32 bit bullseye.
> 
> When booting, this message appears.
> ath9k_htc: Device endpoint numbers are not the expected ones

Did you try searching on that exact language? That often produces
useful results.

https://duckduckgo.com/?q=ath9k_htc%3A+Device+endpoint+numbers+are+not+the+expected+ones

-- 
Does anybody read signatures any more?

https://charlescurley.com
https://charlescurley.com/blog/

TP Link TL-WN722N adapter & ath9k-htc driver.

[peter]

Hi,

An old Sharp Mebius here has 32 bit bullseye.

When booting, this message appears.
ath9k_htc: Device endpoint numbers are not the expected ones

In a console, these.

$ lsusb
Bus 001 Device 002: ID 0cf3:9271 Qualcomm Atheros Communications AR9271 802.11n
Bus 001 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub

$ ip link show
1: lo:  mtu 65536 qdisc noqueue state UNKNOWN mode 
DEFAULT group debault qlen 1000
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
2: enp0s18:  mtu 1500 qdisc pfifo_fast state 
UP mode DEFAULT group default qlen 1000
link/ether 00:40:d0:1b:01:6a brd ff:ff:ff:ff:ff:ff

$ lsmod | grep ath9k
ath9k_htc  61440  0
ath9k_common   16384  1 ath9k_htc
ath9k_hw  421888  2 ath9k_htc,ath9k_common

Can anyone explain "Device endpoint numbers are not the expected ones"?
Other ideas?

Thanks, ... P.

- 
mobile: +1 778 951 5147
VoIP:   +1 604 670 0140

Re: SOLUTION AW: EPSON ET M 1120 new printer: If You can read this, you are using the wrong driver

[Andrew M.A. Cater]

On Wed, May 17, 2023 at 06:27:55PM +, Schwibinger Michael wrote:
> Good evening
> 
> This did work.
> 
> Thank You
> 
> Thank You
> 
> Thank You
> 
> 
> Thank You
> 
> Thank You
> 
> 
Hi Sophie,

I'm really very pleased that it all worked for you eventually.

There were some false starts and some misunderstandings but the main 
thing is that IT WORKS :)

We don't normally delete emails from the mailing lists - they are there
to explain how things happened and how solutions were found.

These emails also led to people explaining various facts about
how to use the mailing list - maybe these will help other people.

Herzlichen Gluckwuenshen

Andy
> 
> 
> 
> Regards Sope
> 
> I ll send 2nd email with topoic   Delete Printer Emails.
> 
> 
> 
> 
> Von: Jeremy Ardley 
> Gesendet: Montag, 8. Mai 2023 00:47
> An: debian-user@lists.debian.org 
> Betreff: Re: EPSON ET M 1120 new printer: If You can read this, you are using 
> the wrong driver
> 
> 
> On 8/5/23 08:12, Will Mengarini wrote:
> > * Brian  [23-05/08=Mo 00:27 +0100]:
> >>
> >> https://download3.ebz.epson.net/dsc/f/03/00/14/48/15/1d37501ad39bd2b5753 \
> >> cce3b2715b3e2fef557/epson-inkjet-printer-escpr_1.7.26-1lsb3.2_amd64.deb
> > That includes a literal space in the middle of that hash
> > (because the space before the backslash is taken literally).
> >
> > However, when I removed that space by hand, I still got "not found":
> 
> The driver is already included in the standard debian distribution
> 
> sudo apt-get install printer-driver-escpr
> 
> Then the usual cups administration to attach driver to printer using lpadmin 
> or http://localhost:631
> 
> 
> --
> Jeremy
> (Lists)
> 

SOLUTION AW: EPSON ET M 1120 new printer: If You can read this, you are using the wrong driver

[Schwibinger Michael]

Good evening

This did work.

Thank You

Thank You

Thank You


Thank You

Thank You





Regards Sope

I ll send 2nd email with topoic   Delete Printer Emails.




Von: Jeremy Ardley 
Gesendet: Montag, 8. Mai 2023 00:47
An: debian-user@lists.debian.org 
Betreff: Re: EPSON ET M 1120 new printer: If You can read this, you are using 
the wrong driver


On 8/5/23 08:12, Will Mengarini wrote:
> * Brian  [23-05/08=Mo 00:27 +0100]:
>>https://download3.ebz.epson.net/dsc/f/03/00/14/48/15/1d37501ad39bd2b5753 \
>> cce3b2715b3e2fef557/epson-inkjet-printer-escpr_1.7.26-1lsb3.2_amd64.deb
> That includes a literal space in the middle of that hash
> (because the space before the backslash is taken literally).
>
> However, when I removed that space by hand, I still got "not found":

The driver is already included in the standard debian distribution

sudo apt-get install printer-driver-escpr

Then the usual cups administration to attach driver to printer using lpadmin or 
http://localhost:631


--
Jeremy
(Lists)

Re: EPSON ET M 1120 new printer: If You can read this, youareusingthe wrong driver

[gene heskett]

On 5/16/23 12:07, Brian wrote:

On Tue 16 May 2023 at 11:24:05 -0400, gene heskett wrote:


On 5/16/23 10:35, Brian wrote:

On Tue 16 May 2023 at 09:58:45 -0400, gene heskett wrote:


On 5/16/23 08:29, Andrew M.A. Cater wrote:


[...]


But yes, this is going nowhere - and there is no lsb package in Debian, I think.

Andy

.

Couple that with my recent discovery that debian seems to be shipping the
cups from apple, last updated by its author in 2019.


Did your investigations involve reading a changelog?


whazzat? Trolling thru /usr/share/ looking for those is a pita. And far more
often, a waste of time, not containing a useful amount of data.


Mmm. That's a novel way to dispose of the usefulness of changelogs and not
have to acknowledge what appears to be disinformation.
  

Given that clue, I would be surprised to find a "cups" driver in the debian
repo's capable of driving a newer epson product.


Please give yourself a treat and use 'lpinfo -m' :).


Thanks, I just did and it appears to show both brothers installed drivers
and the driverless, but driverless is not complete for either printer, where
the makers ppd is. for the laser, toner would last forever, very thin,
faint, had to read output, the mfcj6920dw only uses top tray, which is $8
for 50 sheets glossy photo paper. Tray 2, the bottom one has up to 350 pages
of duplex copy paper, the obvious choice, but the driverless driver doesn't
use it.


You made a comment on Epson products and Debian. I was rather hoping you would
address that rather than bringing up your concern with a possibly buggy printer
from another vendor.

Sorry, I somewhat subscribe to the suggestion Sophie is a chatbot 
learning, and was more concerned with my own problem cups children. But 
a new kernel on this machine, the "server" seems to have fixed the 
armbian bullseye machines.  epson, given their recent attitude about 
linux, and I as a potential customer pretty much quit worrying about 
them since ink or toner for the brothers cuts my per page costs 
noticeably. I've killed a lot of trees since the early 80's. In a way, 
I'm glad they wouldn't sell me a new high voltage board, the color 
reproduction of this inkjet is far far superior to anything I ever got 
out of the color laser.


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, 1940)
If we desire respect for the law, we must first make the law respectable.
 - Louis D. Brandeis
Genes Web page <http://geneslinuxbox.net:6309/>

Re: EPSON ET M 1120 new printer: If You can read this, you areusingthe wrong driver

[Brian]

On Tue 16 May 2023 at 11:24:05 -0400, gene heskett wrote:

> On 5/16/23 10:35, Brian wrote:
> > On Tue 16 May 2023 at 09:58:45 -0400, gene heskett wrote:
> > 
> > > On 5/16/23 08:29, Andrew M.A. Cater wrote:
> > 
> > [...]
> > 
> > > > But yes, this is going nowhere - and there is no lsb package in Debian, 
> > > > I think.
> > > > 
> > > > Andy
> > > > 
> > > > .
> > > Couple that with my recent discovery that debian seems to be shipping the
> > > cups from apple, last updated by its author in 2019.
> > 
> > Did your investigations involve reading a changelog?
> 
> whazzat? Trolling thru /usr/share/ looking for those is a pita. And far more
> often, a waste of time, not containing a useful amount of data.

Mmm. That's a novel way to dispose of the usefulness of changelogs and not
have to acknowledge what appears to be disinformation. 
 
> > > Given that clue, I would be surprised to find a "cups" driver in the 
> > > debian
> > > repo's capable of driving a newer epson product.
> > 
> > Please give yourself a treat and use 'lpinfo -m' :).
> > 
> Thanks, I just did and it appears to show both brothers installed drivers
> and the driverless, but driverless is not complete for either printer, where
> the makers ppd is. for the laser, toner would last forever, very thin,
> faint, had to read output, the mfcj6920dw only uses top tray, which is $8
> for 50 sheets glossy photo paper. Tray 2, the bottom one has up to 350 pages
> of duplex copy paper, the obvious choice, but the driverless driver doesn't
> use it.

You made a comment on Epson products and Debian. I was rather hoping you would
address that rather than bringing up your concern with a possibly buggy printer
from another vendor.

-- 
Brian.

Re: EPSON ET M 1120 new printer: If You can read this,youareusingthe wrong driver

[gene heskett]

On 5/16/23 11:30, Greg Wooledge wrote:

On Tue, May 16, 2023 at 10:52:28AM -0400, gene heskett wrote:

gnutls28-dev worked, all tests pass. thanks Greg. How do I now make
installable deb's? I think I have dh installed.


I wouldn't even try.  That would be a *project*.


Chuckle, but sir, I am doing it, for linuxcnc n an rpi4b! I can install 
them over the top of the buildbots output if I want, or if the buildbot 
gets a tummy ache.



This is the sort of thing that you install in /opt or /usr/local and
write your own systemd unit files or init.d scripts for, possibly
cannibalizing the Debian ones if they exist.


If it works, I'd prefer the packaging system knows about it.


That said, since this involves printers, I would just ask Brian what
to do.  He seems to know them extremely well, certainly better than I
do, though that's a low bar to pass.

.

Thanks Greg.

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, 1940)
If we desire respect for the law, we must first make the law respectable.
 - Louis D. Brandeis
Genes Web page 

Re: EPSON ET M 1120 new printer: If You can read this, youareusingthe wrong driver

[Greg Wooledge]

On Tue, May 16, 2023 at 10:52:28AM -0400, gene heskett wrote:
> gnutls28-dev worked, all tests pass. thanks Greg. How do I now make
> installable deb's? I think I have dh installed.

I wouldn't even try.  That would be a *project*.

This is the sort of thing that you install in /opt or /usr/local and
write your own systemd unit files or init.d scripts for, possibly
cannibalizing the Debian ones if they exist.

That said, since this involves printers, I would just ask Brian what
to do.  He seems to know them extremely well, certainly better than I
do, though that's a low bar to pass.

Re: EPSON ET M 1120 new printer: If You can read this, you areusingthe wrong driver

[gene heskett]

On 5/16/23 10:35, Brian wrote:

On Tue 16 May 2023 at 09:58:45 -0400, gene heskett wrote:


On 5/16/23 08:29, Andrew M.A. Cater wrote:


[...]


But yes, this is going nowhere - and there is no lsb package in Debian, I think.

Andy

.

Couple that with my recent discovery that debian seems to be shipping the
cups from apple, last updated by its author in 2019.


Did your investigations involve reading a changelog?


whazzat? Trolling thru /usr/share/ looking for those is a pita. And far 
more often, a waste of time, not containing a useful amount of data.



Given that clue, I would be surprised to find a "cups" driver in the debian
repo's capable of driving a newer epson product.


Please give yourself a treat and use 'lpinfo -m' :).

Thanks, I just did and it appears to show both brothers installed 
drivers and the driverless, but driverless is not complete for either 
printer, where the makers ppd is. for the laser, toner would last 
forever, very thin, faint, had to read output, the mfcj6920dw only uses 
top tray, which is $8 for 50 sheets glossy photo paper. Tray 2, the 
bottom one has up to 350 pages of duplex copy paper, the obvious choice, 
but the driverless driver doesn't use it.


Thanks Brian.

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, 1940)
If we desire respect for the law, we must first make the law respectable.
 - Louis D. Brandeis
Genes Web page <http://geneslinuxbox.net:6309/>

Re: EPSON ET M 1120 new printer: If You can read this, youareusingthe wrong driver

[gene heskett]

On 5/16/23 10:28, Greg Wooledge wrote:

On Tue, May 16, 2023 at 10:20:38AM -0400, gene heskett wrote:

On 5/16/23 10:03, Jeffrey Walton wrote:

On Tue, May 16, 2023 at 9:59 AM gene heskett  wrote:

Couple that with my recent discovery that debian seems to be shipping
the cups from apple, last updated by its author in 2019. The new cups
site, openprinting.org copy's src cups-master.zip is about 2.5 times the
size of the debian src, and it builds on bullseye with only one option
to "./configure with-tls=no" (presumably because debian doesn't have
libtls, I couldn't find it)


libtls is part of libressl from the OpenBSD folks. See, for example,
https://github.com/bob-beck/libtls/blob/master/TUTORIAL.md .



Thanks Jeff, but it does not appear in synaptic here on a bullseye install.
should I add another repo?  That's scary. I only do that on my buster
machines to add the linuxcnc buildbot because I run master on my machines.


Debian provides OpenSSL (libssl-dev or the older libssl1.0-dev) and GnuTLS
(libgnutls28-dev).  Chances are, at least one of these will work.

gnutls28-dev worked, all tests pass. thanks Greg. How do I now make 
installable deb's? I think I have dh installed.



Most projects that want TLS support are developed around OpenSSL, so
you should try that first.  Or, read the project's INSTALL or README
files and see what it says to use.

.


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, 1940)
If we desire respect for the law, we must first make the law respectable.
 - Louis D. Brandeis
Genes Web page 

Re: EPSON ET M 1120 new printer: If You can read this, you are usingthe wrong driver

[Brian]

On Tue 16 May 2023 at 09:58:45 -0400, gene heskett wrote:

> On 5/16/23 08:29, Andrew M.A. Cater wrote:

[...]

> > But yes, this is going nowhere - and there is no lsb package in Debian, I 
> > think.
> > 
> > Andy
> > 
> > .
> Couple that with my recent discovery that debian seems to be shipping the
> cups from apple, last updated by its author in 2019.

Did your investigations involve reading a changelog?
 
> Given that clue, I would be surprised to find a "cups" driver in the debian
> repo's capable of driving a newer epson product.

Please give yourself a treat and use 'lpinfo -m' :).

-- 
Brian.

Re: EPSON ET M 1120 new printer: If You can read this, you areusingthe wrong driver

[Greg Wooledge]

On Tue, May 16, 2023 at 10:20:38AM -0400, gene heskett wrote:
> On 5/16/23 10:03, Jeffrey Walton wrote:
> > On Tue, May 16, 2023 at 9:59 AM gene heskett  wrote:
> > > Couple that with my recent discovery that debian seems to be shipping
> > > the cups from apple, last updated by its author in 2019. The new cups
> > > site, openprinting.org copy's src cups-master.zip is about 2.5 times the
> > > size of the debian src, and it builds on bullseye with only one option
> > > to "./configure with-tls=no" (presumably because debian doesn't have
> > > libtls, I couldn't find it)
> > 
> > libtls is part of libressl from the OpenBSD folks. See, for example,
> > https://github.com/bob-beck/libtls/blob/master/TUTORIAL.md .

> Thanks Jeff, but it does not appear in synaptic here on a bullseye install.
> should I add another repo?  That's scary. I only do that on my buster
> machines to add the linuxcnc buildbot because I run master on my machines.

Debian provides OpenSSL (libssl-dev or the older libssl1.0-dev) and GnuTLS
(libgnutls28-dev).  Chances are, at least one of these will work.

Most projects that want TLS support are developed around OpenSSL, so
you should try that first.  Or, read the project's INSTALL or README
files and see what it says to use.

Re: EPSON ET M 1120 new printer: If You can read this, you areusingthe wrong driver

[gene heskett]

On 5/16/23 10:03, Jeffrey Walton wrote:

On Tue, May 16, 2023 at 9:59 AM gene heskett  wrote:

[...]
Couple that with my recent discovery that debian seems to be shipping
the cups from apple, last updated by its author in 2019. The new cups
site, openprinting.org copy's src cups-master.zip is about 2.5 times the
size of the debian src, and it builds on bullseye with only one option
to "./configure with-tls=no" (presumably because debian doesn't have
libtls, I couldn't find it)


libtls is part of libressl from the OpenBSD folks. See, for example,
https://github.com/bob-beck/libtls/blob/master/TUTORIAL.md .

I've never used it, so I can't comment on it. (I have used OpenSSL a
lot, but not the OpenBSD port).

Jeff

.
Thanks Jeff, but it does not appear in synaptic here on a bullseye 
install. should I add another repo?  That's scary. I only do that on my 
buster machines to add the linuxcnc buildbot because I run master on my 
machines.


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, 1940)
If we desire respect for the law, we must first make the law respectable.
 - Louis D. Brandeis
Genes Web page 

Re: EPSON ET M 1120 new printer: If You can read this, you are usingthe wrong driver

[Jeffrey Walton]

On Tue, May 16, 2023 at 9:59 AM gene heskett  wrote:
> [...]
> Couple that with my recent discovery that debian seems to be shipping
> the cups from apple, last updated by its author in 2019. The new cups
> site, openprinting.org copy's src cups-master.zip is about 2.5 times the
> size of the debian src, and it builds on bullseye with only one option
> to "./configure with-tls=no" (presumably because debian doesn't have
> libtls, I couldn't find it)

libtls is part of libressl from the OpenBSD folks. See, for example,
https://github.com/bob-beck/libtls/blob/master/TUTORIAL.md .

I've never used it, so I can't comment on it. (I have used OpenSSL a
lot, but not the OpenBSD port).

Jeff

Re: EPSON ET M 1120 new printer: If You can read this, you are usingthe wrong driver

[gene heskett]

On 5/16/23 08:29, Andrew M.A. Cater wrote:

On Tue, May 16, 2023 at 01:55:08PM +0300, Anssi Saari wrote:

Will Mengarini  writes:


* Brian  [23-05/08=Mo 00:27 +0100]:

   https://download3.ebz.epson.net/dsc/f/03/00/14/48/15/1d37501ad39bd2b5753 \
cce3b2715b3e2fef557/epson-inkjet-printer-escpr_1.7.26-1lsb3.2_amd64.deb


That includes a literal space in the middle of that hash
(because the space before the backslash is taken literally).

However, when I removed that space by hand, I still got "not found":

debian/pts/3 bash3 ~ 17:03 0$HEAD
https://download3.ebz.epson.net/dsc/f/03/00/14/48/15/1d37501ad39bd2b5753cce3b2715b3e2fef557/epson-inkjet-printer-escpr_1.7.26-1lsb3.2_amd64.deb


Which is still missing the <> surrounding the actual link. so its still 
a 403.

I don't understand why you guys even propose installing drivers from
Epson when printer-driver-escpr package is in Debian and should be
correct for the printer in question? Is there some established reason to
think it won't work? Or you just didn't do your homework?

I get that this saga probably never goes anywhere but I don't understand
this advice, especially from Andrew. It seems to run counter to the Debian
principle of using software that's actually packaged and tested by Debian.



If you reread - I suggest using the Debian package of escputil - I just wanted
Sophie to see that there was an Epson driver which might provide status
for ink tanks etc. - that's why I went through the steps of exactly how
to find the driver.

But yes, this is going nowhere - and there is no lsb package in Debian, I think.

Andy

.
Couple that with my recent discovery that debian seems to be shipping 
the cups from apple, last updated by its author in 2019. The new cups 
site, openprinting.org copy's src cups-master.zip is about 2.5 times the 
size of the debian src, and it builds on bullseye with only one option 
to "./configure with-tls=no" (presumably because debian doesn't have 
libtls, I couldn't find it)


Given that clue, I would be surprised to find a "cups" driver in the 
debian repo's capable of driving a newer epson product.


However to update this never ending saga, I rebooted this machine into a 
new security patched kernel yesterday, updating this machine to:


5.10.0-23-amd64 #1 SMP Debian 5.10.179-1 (2023-05-12) x86_64 GNU/Linux

and my bullseye armbian machines have magically gotten well. What 
changed? DIIK.


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, 1940)
If we desire respect for the law, we must first make the law respectable.
 - Louis D. Brandeis
Genes Web page <http://geneslinuxbox.net:6309/>

Re: EPSON ET M 1120 new printer: If You can read this, you are using the wrong driver

[Andrew M.A. Cater]

On Tue, May 16, 2023 at 01:55:08PM +0300, Anssi Saari wrote:
> Will Mengarini  writes:
> 
> > * Brian  [23-05/08=Mo 00:27 +0100]:
> >>   https://download3.ebz.epson.net/dsc/f/03/00/14/48/15/1d37501ad39bd2b5753 
> >> \
> >> cce3b2715b3e2fef557/epson-inkjet-printer-escpr_1.7.26-1lsb3.2_amd64.deb
> >
> > That includes a literal space in the middle of that hash
> > (because the space before the backslash is taken literally).
> >
> > However, when I removed that space by hand, I still got "not found":
> > 
> > debian/pts/3 bash3 ~ 17:03 0$HEAD
> > https://download3.ebz.epson.net/dsc/f/03/00/14/48/15/1d37501ad39bd2b5753cce3b2715b3e2fef557/epson-inkjet-printer-escpr_1.7.26-1lsb3.2_amd64.deb
> 
> I don't understand why you guys even propose installing drivers from
> Epson when printer-driver-escpr package is in Debian and should be
> correct for the printer in question? Is there some established reason to
> think it won't work? Or you just didn't do your homework?
> 
> I get that this saga probably never goes anywhere but I don't understand
> this advice, especially from Andrew. It seems to run counter to the Debian
> principle of using software that's actually packaged and tested by Debian.
>

If you reread - I suggest using the Debian package of escputil - I just wanted
Sophie to see that there was an Epson driver which might provide status
for ink tanks etc. - that's why I went through the steps of exactly how
to find the driver.

But yes, this is going nowhere - and there is no lsb package in Debian, I think.

Andy