firmware-iwlwifi

[Georgi Naplatanov]

Dear Debian community,

what is the problem with firmware-iwlwifi in testing?

The package is outdated but it's important.

Kind regards
Georgi

[Georgi Naplatanov]

On 10/2/22 10:22, Brad Rogers wrote:
> On Sat, 1 Oct 2022 20:37:34 +0300
> Georgi Naplatanov <go...@oles.biz> wrote:
>
> Hello Georgi,

>
>> The package is outdated but it's important.
>

> It's the same version as in sid. VCS has -2 (merely bookkeeping
> updates). Bug tracker shows no blocker bugs and no indication of
> anything newer.
>
> If you have specific problems you're encountering, share them here for
> assistance. If that doesn't work out, file a bug report.
>

Hi Brad,

thanks for the advice.

I have a notebook (Lenovo) with Intel's WiFi and I have minor issues
with it - sometimes works as expected but sometimes I cannot connect to
the notebook at first attempt throw SSH or sometimes latency is high. In
kernel log are visible messages for unsuccessful attempts to load
firmware files.

Anyway, I'll try to use upstream's firmware files for Linux kernel and
if these issues disappear I'll file a bug report.

Kind regards
Georgi

[Charles Curley]

On Sun, 2 Oct 2022 11:49:14 +0300
Georgi Naplatanov <go...@oles.biz> wrote:

> I have a notebook (Lenovo) with Intel's WiFi and I have minor issues
> with it - sometimes works as expected but sometimes I cannot connect
> to the notebook at first attempt throw SSH or sometimes latency is
> high.

Which model Lenovo, and what is the exact wifi device? Please use lspci
to identify the device. For example, on my T520:

root@jhegaala:~# lspci | grep -i wifi
03:00.0 Network controller: Realtek Semiconductor Co., Ltd. RTL8188CE 802.11b/g/n WiFi Adapter (rev 01)
root@jhegaala:~#

You may have to use lspci with no options to identify your wifi device.

I'm not sure what you mean by

> … sometimes I cannot connect to the notebook at first attempt throw

> SSH or sometimes latency is high.

Connecting to the network and using SSH are two distinct operations. Do
you connect to a network, and then check that you have a working network
before you try SSH?

What are you using to connect the laptop to a network? Network Manager?

> In kernel log are visible messages for unsuccessful attempts to
> load firmware files.

Do those correlate with successful or unsuccessful attempts to join a
network? Please show us the exact log entry.

--
Does anybody read signatures any more?

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

[The Wanderer]

On 2022-10-02 at 12:33, Charles Curley wrote:

> On Sun, 2 Oct 2022 11:49:14 +0300
> Georgi Naplatanov <go...@oles.biz> wrote:
>
>> I have a notebook (Lenovo) with Intel's WiFi and I have minor issues
>> with it - sometimes works as expected but sometimes I cannot connect
>> to the notebook at first attempt throw SSH or sometimes latency is
>> high.

> I'm not sure what you mean by
>
>> … sometimes I cannot connect to the notebook at first attempt throw
>> SSH or sometimes latency is high.
>
> Connecting to the network and using SSH are two distinct operations. Do
> you connect to a network, and then check that you have a working network
> before you try SSH?
>
> What are you using to connect the laptop to a network? Network Manager?

I'm not sure, but just in case you missed it: please note that he did
not say that he sometimes cannot connect to the *network*, but that he
sometimes cannot connect to the *notebook* through SSH.

I interpret that (with the rest of the context) to mean that he's
connecting the notebook/laptop to the network via Wi-Fi, and then
connecting *to* that laptop *from* another computer via SSH.

And that the SSH connection is (at least part of) what seems to be
dropping out, which is being interpreted as evidence that there's a
problem with the wireless connection.

--
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

[Charles Curley]

On Sun, 02 Oct 2022 12:41:14 -0400
The Wanderer <wand...@fastmail.fm> wrote:

> And that the SSH connection is (at least part of) what seems to be
> dropping out, which is being interpreted as evidence that there's a
> problem with the wireless connection.

Good point. Let's see if the original poster (OP) clarifies this.

[Georgi Naplatanov]

On 10/2/22 19:33, Charles Curley wrote:
> On Sun, 2 Oct 2022 11:49:14 +0300
> Georgi Naplatanov <go...@oles.biz> wrote:
>
>> I have a notebook (Lenovo) with Intel's WiFi and I have minor issues
>> with it - sometimes works as expected but sometimes I cannot connect
>> to the notebook at first attempt throw SSH or sometimes latency is
>> high.
>
> Which model Lenovo, and what is the exact wifi device? Please use lspci
> to identify the device. For example, on my T520:

Hi Charles and The Wanderer,

Actually I think I found out what was happening:
- for failed connection attempts trough SSH - the notebook sometimes
switches to suspend/sleep mode and turns off its WiFi card and it takes
time WiFi connection to become active again
- for high latency - 100 ms and sometimes even more - this problem is
in my router probably (I installed OpenWrt on it, the issue may be in my
local environment, busy WiFi channel because of neighbours' routers, not
mature WiFi driver, etc.)

Here are answers to your questions.

This is partial output from lshw:


description: Notebook
product: 82H8 (LENOVO_MT_82H8_BU_idea_FM_IdeaPad 3 15ITL6)
vendor: LENOVO
version: IdeaPad 3 15ITL6

>
> root@jhegaala:~# lspci | grep -i wifi
> 03:00.0 Network controller: Realtek Semiconductor Co., Ltd. RTL8188CE 802.11b/g/n WiFi Adapter (rev 01)
> root@jhegaala:~#
>
> You may have to use lspci with no options to identify your wifi device.
>

0000:00:14.3 Network controller: Intel Corporation Wi-Fi 6 AX201 (rev 20)

> I'm not sure what you mean by
>
>> … sometimes I cannot connect to the notebook at first attempt throw
>> SSH or sometimes latency is high.

Well, I'll try to explain.

This is a small network with a router connected to Internet, one
workstation connected to the router through Ethernet and the notebook
connected with its WiFi card to the same router.

When the notebook is started, I log in into KDE and KDE connects the
computer to my wireless network automatically. The notebook has
installed Debian testing (up to date), KDE and KDE applet for Network
Manager. After that I try to connect from workstation (Ethernet
connection) through SSH to the notebook (WiFi).

> Connecting to the network and using SSH are two distinct operations. Do
> you connect to a network, and then check that you have a working network
> before you try SSH?

I don't perform any checks usually because KDE connects the notebook
automatically to my network.

> What are you using to connect the laptop to a network? Network Manager?
>

Network manager through applet in KDE.

>> In kernel log are visible messages for unsuccessful attempts to
>> load firmware files.
> Do those correlate with successful or unsuccessful attempts to join a
> network? Please show us the exact log entry.

These are the entries from kernel log:

[ 3.453655] iwlwifi 0000:00:14.3: firmware: failed to load
iwlwifi-QuZ-a0-hr-b0-72.ucode (-2)
[ 3.453695] firmware_class: See https://wiki.debian.org/Firmware for
information about missing firmware
[ 3.454449] iwlwifi 0000:00:14.3: firmware: failed to load
iwlwifi-QuZ-a0-hr-b0-72.ucode (-2)
[ 3.454488] iwlwifi 0000:00:14.3: Direct firmware load for
iwlwifi-QuZ-a0-hr-b0-72.ucode failed with error -2
[ 3.454518] iwlwifi 0000:00:14.3: firmware: failed to load
iwlwifi-QuZ-a0-hr-b0-71.ucode (-2)
[ 3.454553] iwlwifi 0000:00:14.3: firmware: failed to load
iwlwifi-QuZ-a0-hr-b0-71.ucode (-2)
[ 3.454581] iwlwifi 0000:00:14.3: Direct firmware load for
iwlwifi-QuZ-a0-hr-b0-71.ucode failed with error -2
[ 3.454592] iwlwifi 0000:00:14.3: firmware: failed to load
iwlwifi-QuZ-a0-hr-b0-70.ucode (-2)
[ 3.454624] iwlwifi 0000:00:14.3: firmware: failed to load
iwlwifi-QuZ-a0-hr-b0-70.ucode (-2)
[ 3.454651] iwlwifi 0000:00:14.3: Direct firmware load for
iwlwifi-QuZ-a0-hr-b0-70.ucode failed with error -2
[ 3.454662] iwlwifi 0000:00:14.3: firmware: failed to load
iwlwifi-QuZ-a0-hr-b0-69.ucode (-2)
[ 3.454694] iwlwifi 0000:00:14.3: firmware: failed to load
iwlwifi-QuZ-a0-hr-b0-69.ucode (-2)
[ 3.454720] iwlwifi 0000:00:14.3: Direct firmware load for
iwlwifi-QuZ-a0-hr-b0-69.ucode failed with error -2
[ 3.454730] iwlwifi 0000:00:14.3: firmware: failed to load
iwlwifi-QuZ-a0-hr-b0-68.ucode (-2)
[ 3.454761] iwlwifi 0000:00:14.3: firmware: failed to load
iwlwifi-QuZ-a0-hr-b0-68.ucode (-2)
[ 3.454787] iwlwifi 0000:00:14.3: Direct firmware load for
iwlwifi-QuZ-a0-hr-b0-68.ucode failed with error -2
[ 3.454799] iwlwifi 0000:00:14.3: firmware: failed to load
iwlwifi-QuZ-a0-hr-b0-67.ucode (-2)
[ 3.454831] iwlwifi 0000:00:14.3: firmware: failed to load
iwlwifi-QuZ-a0-hr-b0-67.ucode (-2)
[ 3.454858] iwlwifi 0000:00:14.3: Direct firmware load for
iwlwifi-QuZ-a0-hr-b0-67.ucode failed with error -2
[ 3.454868] iwlwifi 0000:00:14.3: firmware: failed to load
iwlwifi-QuZ-a0-hr-b0-66.ucode (-2)
[ 3.454902] iwlwifi 0000:00:14.3: firmware: failed to load
iwlwifi-QuZ-a0-hr-b0-66.ucode (-2)
[ 3.454929] iwlwifi 0000:00:14.3: Direct firmware load for
iwlwifi-QuZ-a0-hr-b0-66.ucode failed with error -2
[ 3.454939] iwlwifi 0000:00:14.3: firmware: failed to load
iwlwifi-QuZ-a0-hr-b0-65.ucode (-2)
[ 3.454970] iwlwifi 0000:00:14.3: firmware: failed to load
iwlwifi-QuZ-a0-hr-b0-65.ucode (-2)
[ 3.454997] iwlwifi 0000:00:14.3: Direct firmware load for
iwlwifi-QuZ-a0-hr-b0-65.ucode failed with error -2
[ 3.455007] iwlwifi 0000:00:14.3: firmware: failed to load
iwlwifi-QuZ-a0-hr-b0-64.ucode (-2)
[ 3.455902] iwlwifi 0000:00:14.3: firmware: failed to load
iwlwifi-QuZ-a0-hr-b0-64.ucode (-2)
[ 3.458342] iwlwifi 0000:00:14.3: Direct firmware load for
iwlwifi-QuZ-a0-hr-b0-64.ucode failed with error -2

I'm not sure that these messages correlate to my attempts to log in into
the notebook through SSH. The messages appear always in kernel log.

Kind regards
Georgi

[Stefan Monnier]

> These are the entries from kernel log:
>
> [ 3.453655] iwlwifi 0000:00:14.3: firmware: failed to load
> iwlwifi-QuZ-a0-hr-b0-72.ucode (-2)
> [ 3.453695] firmware_class: See https://wiki.debian.org/Firmware for
> information about missing firmware
> [ 3.454449] iwlwifi 0000:00:14.3: firmware: failed to load
> iwlwifi-QuZ-a0-hr-b0-72.ucode (-2)
> [ 3.454488] iwlwifi 0000:00:14.3: Direct firmware load for
> iwlwifi-QuZ-a0-hr-b0-72.ucode failed with error -2

[...]

> [ 3.455007] iwlwifi 0000:00:14.3: firmware: failed to load
> iwlwifi-QuZ-a0-hr-b0-64.ucode (-2)
> [ 3.455902] iwlwifi 0000:00:14.3: firmware: failed to load
> iwlwifi-QuZ-a0-hr-b0-64.ucode (-2)
> [ 3.458342] iwlwifi 0000:00:14.3: Direct firmware load for
> iwlwifi-QuZ-a0-hr-b0-64.ucode failed with error -2

This shows that the kernel tried to load starting from version 72 going
down and failed down to 64. AFAICT the Debian package comes with
version 63 of that firmware, so the above messages were probably followed
by a message indicating success in loading
`iwlwifi-QuZ-a0-hr-b0-63.ucode` (which is why you're able to connect
over wifi ;-)

> I'm not sure that these messages correlate to my attempts to log in into the
> notebook through SSH.

Very unlikely: the timestamps indicate that these messages were emitted
3.45 seconds after the very beginning of the boot

You might want to try and find version 72 of that firmware, put it into
/lib/firmware/, then do an `rmmod+insmod` (or reboot) to force reloading the
firmware and then see if your problems/delays get solved.

If they do, you might like to file a bug report against the
`firmware-iwlwifi` Debian package, e.g. with `reportbug`.


Stefan

[David Wright]

On Mon 03 Oct 2022 at 09:25:15 (+0300), Georgi Naplatanov wrote:
> On 10/2/22 19:33, Charles Curley wrote:
> > On Sun, 2 Oct 2022 11:49:14 +0300
> > Georgi Naplatanov <go...@oles.biz> wrote:
> >
> > > I have a notebook (Lenovo) with Intel's WiFi and I have minor issues
> > > with it - sometimes works as expected but sometimes I cannot connect
> > > to the notebook at first attempt throw SSH or sometimes latency is
> > > high.
> >
> > Which model Lenovo, and what is the exact wifi device? Please use lspci
> > to identify the device. For example, on my T520:
>

> Actually I think I found out what was happening:
> - for failed connection attempts trough SSH - the notebook sometimes
> switches to suspend/sleep mode and turns off its WiFi card and it
> takes time WiFi connection to become active again

I wake up my machines remotely, and then have to unlock (encrypted)
/home, so my unlock script is:

unlock-axis ()
{
until ping -c 1 -W 1 axis | grep 'bytes from'; do
printf '%s' "axis ";
done;
printf '%(%H:%M:%S)T\n' -1 && ssh -X axis -l unlock
}

This is related to my impatience rather than any system fault.

> These are the entries from kernel log:
>
> [ 3.453655] iwlwifi 0000:00:14.3: firmware: failed to load
> iwlwifi-QuZ-a0-hr-b0-72.ucode (-2)
> [ 3.453695] firmware_class: See https://wiki.debian.org/Firmware
> for information about missing firmware
> [ 3.454449] iwlwifi 0000:00:14.3: firmware: failed to load
> iwlwifi-QuZ-a0-hr-b0-72.ucode (-2)
> [ 3.454488] iwlwifi 0000:00:14.3: Direct firmware load for
> iwlwifi-QuZ-a0-hr-b0-72.ucode failed with error -2

[ … counting down to … ]

> iwlwifi-QuZ-a0-hr-b0-64.ucode (-2)
> [ 3.455902] iwlwifi 0000:00:14.3: firmware: failed to load
> iwlwifi-QuZ-a0-hr-b0-64.ucode (-2)
> [ 3.458342] iwlwifi 0000:00:14.3: Direct firmware load for
> iwlwifi-QuZ-a0-hr-b0-64.ucode failed with error -2
>
> I'm not sure that these messages correlate to my attempts to log in
> into the notebook through SSH. The messages appear always in kernel
> log.

So is this followed by an entry along the lines of:

iwlwifi 0000:00:14.3: firmware: direct-loading firmware iwlwifi-QuZ-a0-hr-b0-63.ucode

or does it say something like:

iwlwifi 0000:00:14.3: minimum version required: iwlwifi-QuZ-a0-hr-b0-64
iwlwifi 0000:00:14.3: maximum version supported: iwlwifi-QuZ-a0-hr-b0-72

The former would explain why your wifi is working, and
the latter would make it surprising that it still worked.

You normally only see this countdown in the debian-installer, before
the firmware package gets installed by check-missing-firmware.
What's out of the ordinary is that your hardware would prefer a more
up-to-date version (72) than what's in bookworm/sid.

Cheers,
David.

[Georgi Naplatanov]

Hi David,

actually I have in the log:

[ 3.817724] iwlwifi 0000:00:14.3: loaded firmware version
63.c04f3485.0 QuZ-a0-hr-b0-63.ucode op_mode iwlmvm

I tried to attach full output from dmesg, lshw, and lspci but but my
message didn't get into mail list so I deleted all my attachments and
sent my message again.

Kind regards
Georgi

[Georgi Naplatanov]

On 10/3/22 20:00, Stefan Monnier wrote:
>> These are the entries from kernel log:
>>
>> [ 3.453655] iwlwifi 0000:00:14.3: firmware: failed to load
>> iwlwifi-QuZ-a0-hr-b0-72.ucode (-2)
>> [ 3.453695] firmware_class: See https://wiki.debian.org/Firmware for
>> information about missing firmware
>> [ 3.454449] iwlwifi 0000:00:14.3: firmware: failed to load
>> iwlwifi-QuZ-a0-hr-b0-72.ucode (-2)
>> [ 3.454488] iwlwifi 0000:00:14.3: Direct firmware load for
>> iwlwifi-QuZ-a0-hr-b0-72.ucode failed with error -2
> [...]
>> [ 3.455007] iwlwifi 0000:00:14.3: firmware: failed to load
>> iwlwifi-QuZ-a0-hr-b0-64.ucode (-2)
>> [ 3.455902] iwlwifi 0000:00:14.3: firmware: failed to load
>> iwlwifi-QuZ-a0-hr-b0-64.ucode (-2)
>> [ 3.458342] iwlwifi 0000:00:14.3: Direct firmware load for
>> iwlwifi-QuZ-a0-hr-b0-64.ucode failed with error -2
>
> This shows that the kernel tried to load starting from version 72 going
> down and failed down to 64. AFAICT the Debian package comes with
> version 63 of that firmware, so the above messages were probably followed
> by a message indicating success in loading
> `iwlwifi-QuZ-a0-hr-b0-63.ucode` (which is why you're able to connect
> over wifi ;-)

Ah, thanks for explanation,

I thought that these (different numbers) are for different types of
firmwares and devices. I thought that newer version just replaces the
old one.

>> I'm not sure that these messages correlate to my attempts to log in into the
>> notebook through SSH.
>
> Very unlikely: the timestamps indicate that these messages were emitted
> 3.45 seconds after the very beginning of the boot
>
> You might want to try and find version 72 of that firmware, put it into
> /lib/firmware/, then do an `rmmod+insmod` (or reboot) to force reloading the
> firmware and then see if your problems/delays get solved.
>

https://git.kernel.org/pub/scm/linux/kernel/git/firmware/linux-firmware.git/tree/

it seems that version 72 is not available but I'll try with 71.

Kind regards
Georgi

[Charles Curley]

On Mon, 3 Oct 2022 20:53:46 +0300
Georgi Naplatanov <go...@oles.biz> wrote:

> https://git.kernel.org/pub/scm/linux/kernel/git/firmware/linux-firmware.git/tree/
>
> it seems that version 72 is not available but I'll try with 71.

I wouldn't bother. I agree with Stefan Monnier
<mon...@iro.umontreal.ca>'s observation that these occur within four
seconds of booting. It is quite possible that these occur because the
boot process has not completed, e.g. a necessary file system hasn't yet
been mounted. From what you describe elsewhere, you are able to get a
connection afterwards. This sounds to me like a separate, harmless,
problem not worth your time.

> Actually I think I found out what was happening:
> - for failed connection attempts trough SSH - the notebook
> sometimes switches to suspend/sleep mode and turns off its WiFi card
> and it takes time WiFi connection to become active again

This sounds reasonable. Suspend likely shuts off the wifi receiver, and
upon ending the system may take a while to bring it back up.

I wonder why it's going into suspend mode; perhaps it is doing so
unnecessarily.

> - for high latency - 100 ms and sometimes even more - this problem
> is in my router probably (I installed OpenWrt on it, the issue may be
> in my local environment, busy WiFi channel because of neighbours'
> routers, not mature WiFi driver, etc.)

OpenWRT is supposed to be pretty solid these days (although individual
drivers may be problematic. But the interference from neighbors is
worth investigating. You should be able to get a list of neighboring
APs and their channels, but I don't know how to do that offhand. Or
just try a different channel for a week or so.

I have a WRT router, and find it expedient to reboot it occasionally.

[David]

On Tue, 4 Oct 2022 at 06:27, Charles Curley
<charle...@charlescurley.com> wrote:
> On Mon, 3 Oct 2022 20:53:46 +0300 Georgi Naplatanov <go...@oles.biz> wrote:

>> - for high latency - 100 ms and sometimes even more - this problem
>> is in my router probably (I installed OpenWrt on it, the issue may be
>> in my local environment, busy WiFi channel because of neighbours'
>> routers, not mature WiFi driver, etc.)

> But the interference from neighbors is
> worth investigating. You should be able to get a list of neighboring
> APs and their channels, but I don't know how to do that offhand. Or
> just try a different channel for a week or so.

(responding just to the above, not the whole thread)

If a GUI is available, this tool shows the activity of the WiFi channels
received:
https://packages.debian.org/bullseye/linssid

[David Wright]

On Mon 03 Oct 2022 at 13:27:13 (-0600), Charles Curley wrote:

> On Mon, 3 Oct 2022 20:53:46 +0300 Georgi Naplatanov wrote:
>
> > Actually I think I found out what was happening:
> > - for failed connection attempts trough SSH - the notebook
> > sometimes switches to suspend/sleep mode and turns off its WiFi card
> > and it takes time WiFi connection to become active again
>
> This sounds reasonable. Suspend likely shuts off the wifi receiver, and
> upon ending the system may take a while to bring it back up.

I'm not sure I understand how, if the wifi receiver is turned off,
the laptop receives any packets that are intended to (incidentally)
turn it back on and connect to it.

I don't have much experience with sleep states. When my laptop is bone
idle (open, screen blanked, kbd unlit, no logins, on battery), iwd
still scans the wifi every minute, and ping returns are still instant.
I presume this is the state known as S0.

When I close the lid, it sleeps (I presume it's S3 as there's no swap;
much activity occurs, and is logged, when the lid is reopened). Ping
gets no replies.

S1 and S2 are the mystery ones for me, as I have no idea how to
enter/leave them. I suspect they may be the states where people post
about their (non-builtin) keyboards and mice not waking them up.

> I wonder why it's going into suspend mode; perhaps it is doing so
> unnecessarily.

Is this perhaps something that DEs do, along with screensavers and
screen locking. (I don't know whether the OP installs one; I don't.
Perhaps that's why I've not encountered S1/S2.)

Cheers,
David.

[Darac Marjal]

On 05/10/2022 05:18, David Wright wrote:
> On Mon 03 Oct 2022 at 13:27:13 (-0600), Charles Curley wrote:
>> On Mon, 3 Oct 2022 20:53:46 +0300 Georgi Naplatanov wrote:
>>
>>> Actually I think I found out what was happening:
>>> - for failed connection attempts trough SSH - the notebook
>>> sometimes switches to suspend/sleep mode and turns off its WiFi card
>>> and it takes time WiFi connection to become active again
>> This sounds reasonable. Suspend likely shuts off the wifi receiver, and
>> upon ending the system may take a while to bring it back up.
> I'm not sure I understand how, if the wifi receiver is turned off,
> the laptop receives any packets that are intended to (incidentally)
> turn it back on and connect to it.
>
> I don't have much experience with sleep states. When my laptop is bone
> idle (open, screen blanked, kbd unlit, no logins, on battery), iwd
> still scans the wifi every minute, and ping returns are still instant.
> I presume this is the state known as S0.
>
> When I close the lid, it sleeps (I presume it's S3 as there's no swap;
> much activity occurs, and is logged, when the lid is reopened). Ping
> gets no replies.
>
> S1 and S2 are the mystery ones for me, as I have no idea how to
> enter/leave them. I suspect they may be the states where people post
> about their (non-builtin) keyboards and mice not waking them up.

According to
https://github.com/torvalds/linux/blob/master/Documentation/admin-guide/pm/sleep-states.rst,

S1 = Standby

S3 = Suspend-to-RAM

S4  = Hibernation

These states can be achieved (if your hardware supports it) by executing
"echo standby > /sys/power/state", "echo deep > /sys/power/mem_sleep"
and "echo dik > /sys/power/state" (although you may find a userspace
tool such as systemd or a GUI may be more suitable).

(S2 is very similar to S3. Maybe the only difference is that devices
(e.g. the screen) are powered off. I'm not sure that S2 serves much
practical purpose.)