Chromium under Xfce/bookworm anyone?

[to...@tuxteam.de]

Hi,

it seems I can't start chromium. Pretty straight net install,
Xfce desktop environment (no specialties, just the "normal"
install), then "apt-get install chromium".

Issuing "chromium" in a terminal seems to hang, top shows
four processes running (and eating RAM for breakfast), but
well whithin my 8G (at least RES, VIRT is at 32G.

The whole thing stays like this until I ^C the command. Box
behaves otherwise normally.

Starting firefox (from the launcher or from the command line)
works as I'd expect it to.

Am I doing anything wrong?

Sorry for my lack of specificity, I have no experience with
Chromium.

Cheers
--
t

[Curt]

Maybe related to this bug?

https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1039037

[to...@tuxteam.de]

On Tue, Sep 12, 2023 at 03:25:35PM -0000, Curt wrote:
> On 2023-09-12, <to...@tuxteam.de> <to...@tuxteam.de> wrote:
> >
> > Hi,
> >
> > it seems I can't start chromium. Pretty straight net install,
> > Xfce desktop environment (no specialties, just the "normal"
> > install), then "apt-get install chromium".

[...]

> Maybe related to this bug?
>
> https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1039037

The description seems to fit. Thanks, Curt for the find!

Cheers
--
t

[David]

Similar situation to mine, although specifics vary.
The situation here is with Chromium on a SID laptop, while Chromium on
a Stable desktop is just fine.
Starting from a terminal gives:

Gtk-Message: 06:46:36.954: Failed to load module "appmenu-gtk-module"
libva error: /usr/lib/x86_64-linux-gnu/dri/iHD_drv_video.so init failed
[12968:12968:0913/064637.144984:ERROR:chrome_browser_cloud_management_c
ontroller.cc(163)] Cloud management controller initialization aborted
as CBCM is not enabled.
[12968:12999:0913/064637.896873:ERROR:nss_util.cc(357)] After loading
Root Certs, loaded==false: NSS error code: -8018
[0913/064638.283625:ERROR:elf_dynamic_array_reader.h(64)] tag not found
[13049:1:0100/000000.316639:ERROR:broker_posix.cc(41)] Recvmsg error:
Connection reset by peer (104)
Segmentation fault

It just looked like something that needed the developers to wake up in
the morning, so I've left it for now.
I do get a flash of Chromium on the screen before it dies, so it's
classic seg fault. It just looks like the latest development phase
hasn't been married up too well with the context.
Cheers!

--
`One day, the great European war will come out of some damned foolish
thing in the Balkans'.

-- Otto von Bismarck (1888)

[Celejar]

On Wed, 13 Sep 2023 06:55:40 +1000
David <curmu...@telaman.net.au> wrote:


> Similar situation to mine, although specifics vary.
> The situation here is with Chromium on a SID laptop, while Chromium on
> a Stable desktop is just fine.
> Starting from a terminal gives:
>
> Gtk-Message: 06:46:36.954: Failed to load module "appmenu-gtk-module"
> libva error: /usr/lib/x86_64-linux-gnu/dri/iHD_drv_video.so init failed
> [12968:12968:0913/064637.144984:ERROR:chrome_browser_cloud_management_c
> ontroller.cc(163)] Cloud management controller initialization aborted
> as CBCM is not enabled.
> [12968:12999:0913/064637.896873:ERROR:nss_util.cc(357)] After loading
> Root Certs, loaded==false: NSS error code: -8018
> [0913/064638.283625:ERROR:elf_dynamic_array_reader.h(64)] tag not found
> [13049:1:0100/000000.316639:ERROR:broker_posix.cc(41)] Recvmsg error:
> Connection reset by peer (104)
> Segmentation fault
>
> It just looked like something that needed the developers to wake up in
> the morning, so I've left it for now.
> I do get a flash of Chromium on the screen before it dies, so it's
> classic seg fault. It just looks like the latest development phase
> hasn't been married up too well with the context.
> Cheers!

The OP was on Stable; with Sid, we're hitting this:

https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1051355

One solution until the problem is resolved is to downgrade the Chromium
packages, as per here:

https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1051355#10

(The older Chromium packages are available at snapshot.debian.org.)

--
Celejar

[David]

Yes, thanks for that.
I'll just hang in.
I never rely on just the one browser, having at least two installed.
I suspect the fix will be along in the not too distant future.
Chromium development is fairly active.

[to...@tuxteam.de]

On Wed, Sep 13, 2023 at 06:55:40AM +1000, David wrote:

[...]

> Similar situation to mine, although specifics vary.
> The situation here is with Chromium on a SID laptop, while Chromium on
> a Stable desktop is just fine.
> Starting from a terminal gives:
>
> Gtk-Message: 06:46:36.954: Failed to load module "appmenu-gtk-module"
> libva error: /usr/lib/x86_64-linux-gnu/dri/iHD_drv_video.so init failed
> [12968:12968:0913/064637.144984:ERROR:chrome_browser_cloud_management_c
> ontroller.cc(163)] Cloud management controller initialization aborted
> as CBCM is not enabled.
> [12968:12999:0913/064637.896873:ERROR:nss_util.cc(357)] After loading
> Root Certs, loaded==false: NSS error code: -8018
> [0913/064638.283625:ERROR:elf_dynamic_array_reader.h(64)] tag not found
> [13049:1:0100/000000.316639:ERROR:broker_posix.cc(41)] Recvmsg error:
> Connection reset by peer (104)
> Segmentation fault
>
> It just looked like something that needed the developers to wake up in
> the morning, so I've left it for now.
> I do get a flash of Chromium on the screen before it dies, so it's
> classic seg fault. It just looks like the latest development phase
> hasn't been married up too well with the context.

This seems to be a bit different: in my case, some four Chromium
processes stay around, hogging some RAM (a couple of 100MB) and
a tiny bit of CPU. No flashes, no segfaults.

Cheers
--
t

[to...@tuxteam.de]

On Tue, Sep 12, 2023 at 08:09:48PM -0400, Celejar wrote:

[...]

> The OP was on Stable; with Sid, we're hitting this:

Yes.

> https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1051355

In the OP's case, Curt's proposal

https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1039037

fits things better (#1051355 mentions a segfault, in #1039037 the
processes just hang around)

> One solution until the problem is resolved is to downgrade the Chromium
> packages, as per here:
>
> https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1051355#10
>
> (The older Chromium packages are available at snapshot.debian.org.)

...but downgrading might be an idea. Alas, I might not get rhe
chance to try. $DAYJOB $BOSS wants to have (eek) Windoze on the
machine (I won't even think of Chromium on my private laptop ;-)

Cheers
--
t

[Celejar]

I have no choice - there's at least one important site (of a major
financial institution) that I need that simply doesn't work with
Firefox.

--
Celejar

[debia...@howorth.org.uk]

Celejar <cel...@gmail.com> wrote:

> I have no choice - there's at least one important site (of a major
> financial institution) that I need that simply doesn't work with
> Firefox.

If I had anything to do with any financial institution, major or not,
that was so incompetent that their website wasn't browser agnostic I
would rapidly disengage myself from any involvement with that financial
institution. I care about my money!

Actually, I can generalise that. I don't do business with any
organisation whose website won't work in FF.

[David Wright]

On Wed 13 Sep 2023 at 08:24:27 (-0400), Celejar wrote:

> I have no choice - there's at least one important site (of a major
> financial institution) that I need that simply doesn't work with
> Firefox.

Usual question: what does "doesn't work" mean?

Cheers,
David.

[Celejar]

The site frequently goes into some kind of redirect loop - it keeps
jumping back and forth between "logon.example.com" and
"some.page.example.com", preventing me from accessing the site. I can
get around this, at least sometimes, by simply going to the URL bar and
typing in some other example.com url, but this behavior is really
annoying.

The truth is, however, that although I thought I had fully investigated
the problem, I was wrong. Upon further investigation, it seems that it's
my uBlock Origin settings that are causing the problem - disabling it
entirely (as opposed to just for the pages mentioned above) solves the
problem. Now I just have to dive into the logger and try to figure out
what I need to adjust to get things working.

--
Celejar

[Charlie Gibbs]

On Wed Sep 13 10:22:29 2023 David Wright <deb...@lionunicorn.co.uk> wrote:

> On Wed 13 Sep 2023 at 08:24:27 (-0400), Celejar wrote:
>
>> I have no choice - there's at least one important site (of a major
>> financial institution) that I need that simply doesn't work with
>> Firefox.

Tell me about it. I normally use Seamonkey (a Firefox variant),
but even the latest Firefox didn't work.

> Usual question: what does "doesn't work" mean?

I can do most of my online banking with Seamonkey/Firefox, but when
trying to activate a new credit card I got halfway into a set of
screens that I don't normally access, and things just hung.

I decided I might as well waste as much of the bank's time as they
were wasting of mine, so I went down to my local branch and managed
to get a supervisor. The first thing he did was ask me which
browser I was using. I told him I was using Firefox, and he said,
"Never heard of it." (!)

He turned around his terminal so I could get at it, and I went through
the same steps as I had tried at home. Naturally, since it was running
their anointed browser (Microsoft Edge), everything went smoothly.
I pointed out that this is a form of discrimination, and left the bank
muttering things about digital racism. Since the supervisor wasn't
Caucasian, this had a gratifying effect.

--
/~\ Charlie Gibbs | They offer a huge range of
\ / <cgi...@kltpzyxm.invalid> | world-class vulnerabilities
X I'm really at ac.dekanfrus | that only Microsoft can provide.
/ \ if you read it the right way. | -- druck <ne...@druck.org.uk>

[to...@tuxteam.de]

On Wed, Sep 13, 2023 at 06:35:12AM +0200, to...@tuxteam.de wrote:
> On Tue, Sep 12, 2023 at 08:09:48PM -0400, Celejar wrote:
>
> [...]
>
> > The OP was on Stable; with Sid, we're hitting this:
>
> Yes.
>
> > https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1051355
>
> In the OP's case, Curt's proposal
>
> https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1039037

Yes, it is exactly this. After one reboot [1] all is well.

Thank you all for your help!

Cheers

[1] https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1039037#10
--
tomás

[to...@tuxteam.de]

On Wed, Sep 13, 2023 at 01:14:34PM -0400, Celejar wrote:
> On Wed, 13 Sep 2023 09:16:15 -0500
> David Wright <deb...@lionunicorn.co.uk> wrote:
>
> > On Wed 13 Sep 2023 at 08:24:27 (-0400), Celejar wrote:
> >
> > > I have no choice - there's at least one important site (of a major
> > > financial institution) that I need that simply doesn't work with
> > > Firefox.
> >
> > Usual question: what does "doesn't work" mean?
>
> The site frequently goes into some kind of redirect loop - it keeps
> jumping back and forth between "logon.example.com" and
> "some.page.example.com",

This looks like some cookie gets lost in transaction. The second
page sees you coming empty-handed, sends you to the first to get
your cookie and for some reason you drop it (I know, because I
usually disallow all cookies ;-)

> preventing me from accessing the site. I can
> get around this, at least sometimes, by simply going to the URL bar and
> typing in some other example.com url, but this behavior is really
> annoying.
>
> The truth is, however, that although I thought I had fully investigated
> the problem, I was wrong. Upon further investigation, it seems that it's
> my uBlock Origin settings that are causing the problem - disabling it
> entirely (as opposed to just for the pages mentioned above) solves the
> problem. Now I just have to dive into the logger and try to figure out
> what I need to adjust to get things working.

Yep, exactly that. It might also happen if you disallow third-party
cookies (a reasonable thing to do) and the bank's web "developers"
play cookie ping-pong with some other shady web "service" (to them,
not to you).

Complain to them. It may not help upfront but it binds resources. If
many people do, they might ending up giving in.

Cheers
--
t

[Michael Kjörling]

On 13 Sep 2023 13:14 -0400, from cel...@gmail.com (Celejar):

> The truth is, however, that although I thought I had fully investigated
> the problem, I was wrong. Upon further investigation, it seems that it's
> my uBlock Origin settings that are causing the problem - disabling it
> entirely (as opposed to just for the pages mentioned above) solves the
> problem. Now I just have to dive into the logger and try to figure out
> what I need to adjust to get things working.

For those times when I just want something to work _once_ and don't
foresee a need to do the thing repeatedly, I have written a script to
start Firefox with a brand new profile, then delete it on exit.

It's not a great solution, admittedly, and the script itself is a bit
of a hack; but it works.

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

[Celejar]

On Thu, 14 Sep 2023 07:04:50 +0000
Michael Kjörling <2695bd...@ewoof.net> wrote:

> On 13 Sep 2023 13:14 -0400, from cel...@gmail.com (Celejar):
> > The truth is, however, that although I thought I had fully investigated
> > the problem, I was wrong. Upon further investigation, it seems that it's
> > my uBlock Origin settings that are causing the problem - disabling it
> > entirely (as opposed to just for the pages mentioned above) solves the
> > problem. Now I just have to dive into the logger and try to figure out
> > what I need to adjust to get things working.
>
> For those times when I just want something to work _once_ and don't
> foresee a need to do the thing repeatedly, I have written a script to
> start Firefox with a brand new profile, then delete it on exit.
>
> It's not a great solution, admittedly, and the script itself is a bit
> of a hack; but it works.

That's what I just did manually to troubleshoot: create a fresh
profile, verify that the site works with it, then delete it.

--
Celejar