Re: Do not know which package bug is in

[David Christensen]

On 04/01/2016 04:42 AM, to...@tuxteam.de wrote:

Another angle of attack would be: is the box reachable via network?
(e.g. install openssh-server, log in via ssh).


+1


(I always check the "SSH server" box on the "Choose software" page when 
installing Debian.)



Once you're in, then it's a matter of figuring out where to look for 
clues.  I'd start with the 'dmesg' command and then look at the contents 
of the files in /var/log.



David

Re: Do not know which package bug is in

[Adam Wilson]

On Fri, 1 Apr 2016 12:43:51 +0100
Robin Oberg  wrote:

> On Fri, 2016-04-01 at 12:00 +0100, Oliver Elphick wrote:
> > On Fri, 2016-04-01 at 11:34 +0100, Robin Oberg wrote:
> > > On Fri, 2016-04-01 at 11:21 +0100, Oliver Elphick wrote:
> > > > Intermittent failures like that sound more like hardware problems.
> > > > I
> > > > seem to recall reading that charging devices like that demands more
> > > > power than the computer can supply.
> > > > 
> > > > Oliver Elphick
> > 
> > > Would that not mean that the same problem exists in other operating
> > > systems as well? But seeing as it works fine to charge this old
> > > iPhone 4
> > > in Windows, so it does not seem like a hardware malfunction in this
> > > particular case.
> > 
> > Not necessarily. It might be that Windows doesn't use a particular area
> > of memory that Linux does.
> > 
> > I should go for the other poster's suggestion, of using a powered USB
> > hub. If the failures cease, it was a hardware problem.
> > 
> > 
> 
> Of course, unplugging the device from the USB port stops the crashing,
> because the crashing starts when the device is plugged in to begin
> with...
> 
> If "Linux" is programmed to use a particular area of memory that makes
> it crash, then this is a software related issue, isn't it?

Why the "Linux"? He was actually talking about Linux.

And yes, that would make it a software issue. But semantics.


pgpFebx879hgO.pgp
Description: OpenPGP digital signature

Re: Do not know which package bug is in

[Joe]

On Fri, 01 Apr 2016 11:21:27 +0100
Oliver Elphick  wrote:

> On Fri, 2016-04-01 at 10:25 +0100, Robin Oberg wrote:
> > Hi!
> > 
> > I'm new to Linux and Debian 8.3, but thought I'd help out by sharing
> > this bug that I've found. My system crashes randomly when my iPhone
> > is
> > charging via USB, and every time that it happens I am forced to
> > hard-reboot the laptop by holding down the power button. I don't
> > know where in the system this bug could be located, what package it
> > might be.  
> ...
> > 
> > Any help would be appreciated.  
> 
> Intermittent failures like that sound more like hardware problems.

Indeed so, but we have been somewhat conditioned to believe that
catastrophic failure with the CPU(s) either halted or looping
uselessly is inevitably a sign of hardware failure. I would always bet
that way, but sometimes I'd lose.

I once tried producing Bluetooth sound from a Wheezy machine, and
plugging in a dongle killed the dual-core machine stone dead. No
response to network stimuli, no logs even a second past the point of
failure, no error messages at that point, just the expected USB new
connection stuff. I tried a few more times with the same result (don't
quote Einstein at me, he had never seen a modern computer, which is
not determinate on any human scale).

I accept that I must not have had the right software installed to deal
with this dongle, but this was definitely a software-related hard
crash, so it can happen. Mine was instant and utterly repeatable, but it
is even easier for software failure to be intermittent than hardware
failure. Einstein hadn't heard of intermittent faults, either.

I've used the same (motherboard) USB ports for other peripherals,
including charging, though not at iPhone levels, which certainly are
beyond the USB spec. The same dongle has been used to talk to a GPS
device and a CAN bus device while connected to a netbook running sid at
a time roughly contemporary with Wheezy.

-- 
Joe

Re: Do not know which package bug is in

[Felix Miata]

Oliver Elphick composed on 2016-04-01 12:00 (UTC+0100)


On Fri, 2016-04-01 at 11:34 +0100, Robin Oberg wrote:



On Fri, 2016-04-01 at 11:21 +0100, Oliver Elphick wrote:



> Intermittent failures like that sound more like hardware problems. I
> seem to recall reading that charging devices like that demands more
> power than the computer can supply.



Would that not mean that the same problem exists in other operating
systems as well? But seeing as it works fine to charge this old
iPhone 4
in Windows, so it does not seem like a hardware malfunction in this
particular case.



Not necessarily. It might be that Windows doesn't use a particular area
of memory that Linux does.



I should go for the other poster's suggestion, of using a powered USB
hub. If the failures cease, it was a hardware problem.


Maybe only indirectly a hardware problem, handled well enough by Windows 
power management, but not by the Linux kernel's default power management 
configuration.


Robin, does the crashing occur regardless which laptop USB port is used? What 
is the laptop's brand and model?

--
"The wise are known for their understanding, and pleasant
words are persuasive." Proverbs 16:21 (New Living Translation)

 Team OS/2 ** Reg. Linux User #211409 ** a11y rocks!

Felix Miata  ***  http://fm.no-ip.com/

Re: Do not know which package bug is in

[Gene Heskett]

On Friday 01 April 2016 07:43:51 Robin Oberg wrote:

> On Fri, 2016-04-01 at 12:00 +0100, Oliver Elphick wrote:
> > On Fri, 2016-04-01 at 11:34 +0100, Robin Oberg wrote:
> > > On Fri, 2016-04-01 at 11:21 +0100, Oliver Elphick wrote:
> > > > Intermittent failures like that sound more like hardware
> > > > problems. I
> > > > seem to recall reading that charging devices like that demands
> > > > more power than the computer can supply.
> > > >
> > > > Oliver Elphick
> > >
> > > Would that not mean that the same problem exists in other
> > > operating systems as well? But seeing as it works fine to charge
> > > this old iPhone 4
> > > in Windows, so it does not seem like a hardware malfunction in
> > > this particular case.
> >
> > Not necessarily. It might be that Windows doesn't use a particular
> > area of memory that Linux does.
> >
> > I should go for the other poster's suggestion, of using a powered
> > USB hub. If the failures cease, it was a hardware problem.
>
> Of course, unplugging the device from the USB port stops the crashing,
> because the crashing starts when the device is plugged in to begin
> with...
>
> If "Linux" is programmed to use a particular area of memory that makes
> it crash, then this is a software related issue, isn't it?
>
> Isn't there a log file somewhere that can confirm what the problem is,
> whether or not it is hardware related, or at least which package the
> crash is related to?
>
> //Robin Oberg

I believe another poster had the best idea.  Locate another machine and 
plug it into your network, doing the config to put your 2nd machine on 
the same class D as your troublesome one is, and "ssh -Y 
troublesome_machine" from the 2nd one as the same user.  If you are 
first user on both machines its probably a lot easier.

Then, when the lockup occurs, the last screenfull of a freshly 
done "dmesg" o that ligin to the first machines screen should show you 
what happened.

I do that here for 5 machines, all using static entries 
in /etc/network/interfaces file to define that machines address when 
bringing up the port at boot or network restart time, 
My /etc/resolv.conf says "host,dns", and an identical /etrc/hosts file 
on all machines that links the machine FQDN and its shorthand alias to a 
given address.

No dhcp involved, all host file lookup.  DNS is defined in the interfaces 
file as the address of my router, so a hostname lookup failure is 
forwarded to the router, and if it doesn't know, its forwarded to my 
ISP's dns servers.  Gateway is also defined in the interfaces file as 
the routers address. And I never have those protracted waits while dhcp 
works.  And full internet access from any of those machines sure makes 
it handy to keep them up to date.  The router of course is running 
dd-wrt.  Best guard dog ever IMO.

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)
Genes Web page 

Re: Do not know which package bug is in

[Thomas Schmitt]

Hi,

Robin Oberg wrote:
> Isn't there a log file somewhere that can confirm what the problem is,

Try this shell command as superuser or via sudo

  dmesg | less

Look into the files in directory /var/log .
I find device related messages in:

  /var/log/kern.log
  /var/log/messages

and their older records kern.log.1 , messages.1, kern.log.2.gz,
messages.2.gz ...


Have a nice day :)

Thomas

Re: Do not know which package bug is in

[tomas]

-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

On Fri, Apr 01, 2016 at 02:02:23PM +0200, arian wrote:
> hi Robin,
> 
> > Everything except the mouse just freezes, and no keyboard combination I
> > can think of helps. Ctrl+C, Ctrl+X, Ctrl+Q, Ctrl+Esc, Super+Esc, Ctrl
> > +Alt+Del... There is only one way out, and that's to hard-reboot
> > everything (which my harddrive doesn't seem to enjoy).
> 
> try magic sysrequests:
> alt-print-k to kill everything on the current terminal (you should get a new 
> login screen)
> alt-print-r to reset the keyboard configuration, followed by switching to 
> another terminal (ctrl-alt-[f1,f2,...] - you should get a text login screen. 
> Maybe there are kernel messages of the form:
> [ uptime ] message. If there are, you could post a screenshot (you know, one 
> with a camera) of them.

I'd guess that the USB hangs: then sysrequests won't reach the box
anyway, since the keyboard is on the USB (but still, it's worth a
try).

Another angle of attack would be: is the box reachable via network?
(e.g. install openssh-server, log in via ssh).

As for the logs... /var/log/syslog resp /var/log/messages (or wherever
systemd hides those things these days -- there I'd have to defer to
others).

regards
- -- tomás
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.12 (GNU/Linux)

iEYEARECAAYFAlb+XrcACgkQBcgs9XrR2kbDzwCffEPj1+IDnLQU4QzhSfMBQr3W
NHAAnjEQtJZNRmPnlmM2o+oSwRPFGn89
=5dgi
-END PGP SIGNATURE-

Re: Do not know which package bug is in

[arian]

hi Robin,

> Everything except the mouse just freezes, and no keyboard combination I
> can think of helps. Ctrl+C, Ctrl+X, Ctrl+Q, Ctrl+Esc, Super+Esc, Ctrl
> +Alt+Del... There is only one way out, and that's to hard-reboot
> everything (which my harddrive doesn't seem to enjoy).

try magic sysrequests:
alt-print-k to kill everything on the current terminal (you should get a new 
login screen)
alt-print-r to reset the keyboard configuration, followed by switching to 
another terminal (ctrl-alt-[f1,f2,...] - you should get a text login screen. 
Maybe there are kernel messages of the form:
[ uptime ] message. If there are, you could post a screenshot (you know, one 
with a camera) of them.

regards, Arian



signature.asc
Description: OpenPGP digital signature

Re: Do not know which package bug is in

[Robin Oberg]

On Fri, 2016-04-01 at 12:00 +0100, Oliver Elphick wrote:
> On Fri, 2016-04-01 at 11:34 +0100, Robin Oberg wrote:
> > On Fri, 2016-04-01 at 11:21 +0100, Oliver Elphick wrote:
> > > Intermittent failures like that sound more like hardware problems.
> > > I
> > > seem to recall reading that charging devices like that demands more
> > > power than the computer can supply.
> > > 
> > > Oliver Elphick
> 
> > Would that not mean that the same problem exists in other operating
> > systems as well? But seeing as it works fine to charge this old
> > iPhone 4
> > in Windows, so it does not seem like a hardware malfunction in this
> > particular case.
> 
> Not necessarily. It might be that Windows doesn't use a particular area
> of memory that Linux does.
> 
> I should go for the other poster's suggestion, of using a powered USB
> hub. If the failures cease, it was a hardware problem.
> 
> 

Of course, unplugging the device from the USB port stops the crashing,
because the crashing starts when the device is plugged in to begin
with...

If "Linux" is programmed to use a particular area of memory that makes
it crash, then this is a software related issue, isn't it?

Isn't there a log file somewhere that can confirm what the problem is,
whether or not it is hardware related, or at least which package the
crash is related to?

//Robin Oberg

Re: Do not know which package bug is in

[tomas]

-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

On Fri, Apr 01, 2016 at 11:34:52AM +0100, Robin Oberg wrote:
> On Fri, 2016-04-01 at 11:21 +0100, Oliver Elphick wrote:
> > On Fri, 2016-04-01 at 10:25 +0100, Robin Oberg wrote:
> > > Hi!
> > > 
> > > I'm new to Linux and Debian 8.3, but thought I'd help out by sharing
> > > this bug that I've found. My system crashes randomly when my iPhone

[...]

> Everything except the mouse just freezes, and no keyboard combination I
> can think of helps. Ctrl+C, Ctrl+X, Ctrl+Q, Ctrl+Esc, Super+Esc, Ctrl
> +Alt+Del... There is only one way out, and that's to hard-reboot
> everything (which my harddrive doesn't seem to enjoy).

Now that's more interesting: the mouse seems to work, that might mean
that the OS is still alive (but something horrible has happened to the
USB so that your desktop is not responding).

Yes "but it works in Windows" might just mean that Windows initializes
the USB in a slightly different way[1]. It'll be quite difficult to
actually find out without access to your very specific configuration.

[1] Not necessarily standard, mind you.

regards
- -- t
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.12 (GNU/Linux)

iEYEARECAAYFAlb+UzwACgkQBcgs9XrR2ka1IACdF/IlXq/HOXfK1wtW6z41xldn
zPsAnRvawoGWdMZRVfntJ40VNmh60VoB
=8amu
-END PGP SIGNATURE-

Re: Do not know which package bug is in

[Oliver Elphick]

On Fri, 2016-04-01 at 11:34 +0100, Robin Oberg wrote:
> On Fri, 2016-04-01 at 11:21 +0100, Oliver Elphick wrote:
> > Intermittent failures like that sound more like hardware problems.
> > I
> > seem to recall reading that charging devices like that demands more
> > power than the computer can supply.
> > 
> > Oliver Elphick

> Would that not mean that the same problem exists in other operating
> systems as well? But seeing as it works fine to charge this old
> iPhone 4
> in Windows, so it does not seem like a hardware malfunction in this
> particular case.

Not necessarily. It might be that Windows doesn't use a particular area
of memory that Linux does.

I should go for the other poster's suggestion, of using a powered USB
hub. If the failures cease, it was a hardware problem.

Re: Do not know which package bug is in

[Robin Oberg]

On Fri, 2016-04-01 at 11:21 +0100, Oliver Elphick wrote:
> On Fri, 2016-04-01 at 10:25 +0100, Robin Oberg wrote:
> > Hi!
> > 
> > I'm new to Linux and Debian 8.3, but thought I'd help out by sharing
> > this bug that I've found. My system crashes randomly when my iPhone
> > is
> > charging via USB, and every time that it happens I am forced to
> > hard-reboot the laptop by holding down the power button. I don't know
> > where in the system this bug could be located, what package it might
> > be.
> ...
> > 
> > Any help would be appreciated.
> 
> Intermittent failures like that sound more like hardware problems. I
> seem to recall reading that charging devices like that demands more
> power than the computer can supply.
> 
> Oliver Elphick

Would that not mean that the same problem exists in other operating
systems as well? But seeing as it works fine to charge this old iPhone 4
in Windows, so it does not seem like a hardware malfunction in this
particular case.

Everything except the mouse just freezes, and no keyboard combination I
can think of helps. Ctrl+C, Ctrl+X, Ctrl+Q, Ctrl+Esc, Super+Esc, Ctrl
+Alt+Del... There is only one way out, and that's to hard-reboot
everything (which my harddrive doesn't seem to enjoy).

//Robin Oberg

Re: Do not know which package bug is in

[tomas]

-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

On Fri, Apr 01, 2016 at 11:21:27AM +0100, Oliver Elphick wrote:
> On Fri, 2016-04-01 at 10:25 +0100, Robin Oberg wrote:
> > Hi!
> > 
> > I'm new to Linux and Debian 8.3, but thought I'd help out by sharing
> > this bug that I've found. My system crashes randomly when my iPhone
> > is
> > charging via USB, and every time that it happens I am forced to
> > hard-reboot the laptop by holding down the power button. I don't know
> > where in the system this bug could be located, what package it might
> > be.
> ...
> > 
> > Any help would be appreciated.
> 
> Intermittent failures like that sound more like hardware problems. I
> seem to recall reading that charging devices like that demands more
> power than the computer can supply.

Yeah -- it may well be flaky hardware: the USB is specified to survive
that (and shut off the offending device), but who knows.

Try interposing a powered USB hub between your computer and the phone.

regards
- -- t
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.12 (GNU/Linux)

iEYEARECAAYFAlb+RaEACgkQBcgs9XrR2kampACfQiBdsZFuTvLu7XqhchdgbnlJ
fsUAnRlun1WM3szNbdiu6T0AATQSRGCE
=6Pmf
-END PGP SIGNATURE-

Re: Do not know which package bug is in

[Oliver Elphick]

On Fri, 2016-04-01 at 10:25 +0100, Robin Oberg wrote:
> Hi!
> 
> I'm new to Linux and Debian 8.3, but thought I'd help out by sharing
> this bug that I've found. My system crashes randomly when my iPhone
> is
> charging via USB, and every time that it happens I am forced to
> hard-reboot the laptop by holding down the power button. I don't know
> where in the system this bug could be located, what package it might
> be.
...
> 
> Any help would be appreciated.

Intermittent failures like that sound more like hardware problems. I
seem to recall reading that charging devices like that demands more
power than the computer can supply.

Oliver Elphick

Do not know which package bug is in

[Robin Oberg]

Hi!

I'm new to Linux and Debian 8.3, but thought I'd help out by sharing
this bug that I've found. My system crashes randomly when my iPhone is
charging via USB, and every time that it happens I am forced to
hard-reboot the laptop by holding down the power button. I don't know
where in the system this bug could be located, what package it might be.
And that's apparently a required thing to do know when one is reporting
a bug.

Any help would be appreciated.

//Robin Oberg