GNU bug report logs - #30467
GuixSD fails to display login prompt on headless server

Previous Next

To add a comment to this bug, you must first unarchive it, by sending
a message to control AT debbugs.gnu.org, with unarchive 30467 in the body.
You can then email your comments to 30467 AT debbugs.gnu.org in the normal way.

Toggle the display of automated, internal messages from the tracker.

Message #5 received at submit <at> debbugs.gnu.org (full text, mbox):

From: George myglc2 Clemmer <myglc2 <at> gmail.com>
To: bug-guix <at> gnu.org
Subject: GuixSD fails to display login prompt on headless server
Date: Thu, 15 Feb 2018 10:01:24 -0500

[Message part 1 (text/plain, inline)]

GuixSD fails to display login prompt on headless server

This was working in (GNU Guix) 0.13.0.4589-74bea6
ae0307f7c tests: Adjust to new unbound-variable error message.

It fails in guix (GNU Guix) 0.13.0.5042-6e385.
6e385b76e gnu: mongodb: Use scons-build-system.

System Generation on g1:
     19            GNU with Linux-Libre 4.14.2 (beta)           61 2017-11-30 19:18:15
     18            GNU with Linux-Libre 4.13.12 (beta)          62 2017-11-11 13:05:56

I just noticed this or I would have reported sooner.

Can check/test on either generation if it would be of help.

TIA - George

config:

[sys.scm (application/octet-stream, attachment)]

Message #8 received at 30467 <at> debbugs.gnu.org (full text, mbox):

From: Danny Milosavljevic <dannym <at> scratchpost.org>
To: George myglc2 Clemmer <myglc2 <at> gmail.com>
Cc: 30467 <at> debbugs.gnu.org
Subject: Re: bug#30467: GuixSD fails to display login prompt on headless server
Date: Thu, 15 Feb 2018 16:54:55 +0100

Hi George,

Leo added a comment to the services/base.scm code:

;;; FIXME This doesn't work as expected. According to agetty(8), if this option
;;; is not passed, then the default is 'auto'. However, in my tests, when that
;;; option is selected, agetty never presents the login prompt, and the
;;; term-ttyS0 service respawns every few seconds.
                                #$@(if local-line
                                       #~(#$(match local-line
                                              ('auto "--local-line=auto")
                                              ('always "--local-line=always")
                                              ('never "-local-line=never")))
                                       #~())

So try specifying local-line ?



Also, can you try updating again?

I've pushed some agetty consolidation to guix-master (only) today.

Also, you have

 (kernel-arguments '("console=ttyS1,115200"))

AND

(agetty-service (agetty-configuration (tty "ttyS1")
							 (baud-rate "115200")))

.

Starting today, the second one will be automatically configured by %base-services.

I wonder whether we can provide a nice upgrade path for users.

Can be automatically remove duplicates from the service definitions?
Can we warn about duplicates in the final service definition?

Does agetty lock the serial port?
Or does it happily start two gettys?  The latter would be bad.

So please keep your config the same and reconfigure and let's see what happens.

In the long run please remove your agetty-service, it should automatically appear
(it recovers the settings from the "console" kernel-argument).

If it doesn't work you can always return to the previous generation so no worries!

Message #11 received at 30467 <at> debbugs.gnu.org (full text, mbox):

From: Leo Famulari <leo <at> famulari.name>
To: Danny Milosavljevic <dannym <at> scratchpost.org>
Cc: 30467 <at> debbugs.gnu.org, George myglc2 Clemmer <myglc2 <at> gmail.com>
Subject: Re: bug#30467: GuixSD fails to display login prompt on headless server
Date: Thu, 15 Feb 2018 11:23:10 -0500

[Message part 1 (text/plain, inline)]

On Thu, Feb 15, 2018 at 04:54:55PM +0100, Danny Milosavljevic wrote:
> Leo added a comment to the services/base.scm code:
> 
> ;;; FIXME This doesn't work as expected. According to agetty(8), if this option
> ;;; is not passed, then the default is 'auto'. However, in my tests, when that
> ;;; option is selected, agetty never presents the login prompt, and the
> ;;; term-ttyS0 service respawns every few seconds.
>                                 #$@(if local-line
>                                        #~(#$(match local-line
>                                               ('auto "--local-line=auto")
>                                               ('always "--local-line=always")
>                                               ('never "-local-line=never")))
>                                        #~())
> 
> So try specifying local-line ?

To clarify, I meant that the documented agetty default (auto) works, but
actually specifying 'auto' in config.scm would break agetty.

So, normally I recommend ignoring this option, since it seems broken
upstream.

[signature.asc (application/pgp-signature, inline)]

Message #14 received at 30467 <at> debbugs.gnu.org (full text, mbox):

From: George myglc2 Clemmer <myglc2 <at> gmail.com>
To: Danny Milosavljevic <dannym <at> scratchpost.org>
Cc: 30467 <at> debbugs.gnu.org, Leo Famulari <leo <at> famulari.name>
Subject: Re: bug#30467: GuixSD fails to display login prompt on headless server
Date: Thu, 15 Feb 2018 16:29:11 -0500

[Message part 1 (text/plain, inline)]

Hi Danny & Leo,

Thanks for the quick response!

On 02/15/2018 at 15:54 Danny Milosavljevic writes:

> Hi George,
>
> Leo added a comment to the services/base.scm code:
>
> ;;; FIXME This doesn't work as expected. According to agetty(8), if this option
> ;;; is not passed, then the default is 'auto'. However, in my tests, when that
> ;;; option is selected, agetty never presents the login prompt, and the
> ;;; term-ttyS0 service respawns every few seconds.
>                                 #$@(if local-line
>                                        #~(#$(match local-line
>                                               ('auto "--local-line=auto")
>                                               ('always "--local-line=always")
>                                               ('never "-local-line=never")))
>                                        #~())
>
> So try specifying local-line ?

Not clear what you wanted so I didn't try this.

> Also, can you try updating again?
>
> I've pushed some agetty consolidation to guix-master (only) today.

I did that, see below ...

> Also, you have
>
>  (kernel-arguments '("console=ttyS1,115200"))
>
> AND
>
> (agetty-service (agetty-configuration (tty "ttyS1")
> 							 (baud-rate "115200")))
> Starting today, the second one will be automatically configured by
> %base-services.

I removed agetty-service but agetty didn't run, see below ...

> So please keep your config the same and reconfigure and let's see what
> happens.

I did git-pull & make to get guix (GNU Guix) 0.14.0.1879-06a9 and 'guix
system reconfigure' >>> the console is still blank.  (my G36)

> In the long run please remove your agetty-service, it should automatically appear
> (it recovers the settings from the "console" kernel-argument).

Then I removed agetty-service, reconfigured and rebooted >>> the console
is still blank. (my G37, sys.scm attached)

FWIW, I don't see agetty running ...

g1 <at> g1 ~$ pstree -a -A
shepherd --no-auto-compile/gnu/store/hjfwjz9g0jrjcyggybi
  |-dhclient -nw -pf /var/run/dhclient.pid enp4s0enp3
  |-guix-daemon --build-users-group guixbuild--
  |-mingetty --noclear tty6
  |-mingetty --noclear tty5
  |-mingetty --noclear tty4
  |-mingetty --noclear tty3
  |-mingetty --noclear tty2
  |-mingetty --noclear tty1
  |-nscd -f/gnu/store/gh3xhwgh927md5ifzdmg0cgpqc0as1sh-ns
  |   `-7*[{nscd}]
  |-ntpd -n -c/gnu/store/1l1yf5dss8r9pqxklaax32s6bkah09c
  |   `-{ntpd}
  |-sshd -D -f/gnu/store/7z8qc04s48cgzvpc77715r3ckh7qr
  [...]
  |-syslogd --rcfile/gnu/store/y5nrfbj52vlnj77iyk
  |-udevd
  `-8*[{shepherd}]

TIA - George

[sys.scm (application/octet-stream, attachment)]

Message #17 received at 30467 <at> debbugs.gnu.org (full text, mbox):

From: George myglc2 Clemmer <myglc2 <at> gmail.com>
To: Danny Milosavljevic <dannym <at> scratchpost.org>
Cc: 30467 <at> debbugs.gnu.org, Leo Famulari <leo <at> famulari.name>
Subject: Re: bug#30467: GuixSD fails to display login prompt on headless server
Date: Thu, 15 Feb 2018 19:24:22 -0500

On 02/15/2018 at 21:29 George myglc writes:

> Hi Danny & Leo,
>
> Thanks for the quick response!
>
> On 02/15/2018 at 15:54 Danny Milosavljevic writes:
[...]
>> Also, you have
>>
>>  (kernel-arguments '("console=ttyS1,115200"))
>>
>> AND
>>
>> (agetty-service (agetty-configuration (tty "ttyS1")
>> 							 (baud-rate "115200")))
>> Starting today, the second one will be automatically configured by
>> %base-services.

> I did git-pull & make to get guix (GNU Guix) 0.14.0.1879-06a9 and 'guix
> system reconfigure' >>> the console is still blank.  (my G36)
>
>> In the long run please remove your agetty-service, it should automatically appear
>> (it recovers the settings from the "console" kernel-argument).
>
> Then I removed agetty-service, reconfigured and rebooted >>> the console
> is still blank. (my G37, sys.scm attached)

BTW, IPMI serial over lan (SOL) which most recently worked
here on ...
: a48b877e6 gnu: appstream-glib: Propagate some inputs.
: guix (GNU Guix) 0.14.0.1505-a48b8

... is also now broken, with or without ...

>> (agetty-service (agetty-configuration (tty "ttyS1")
>> 							 (baud-rate "115200")))

Message #20 received at 30467 <at> debbugs.gnu.org (full text, mbox):

From: Danny Milosavljevic <dannym <at> scratchpost.org>
To: George myglc2 Clemmer <myglc2 <at> gmail.com>
Cc: 30467 <at> debbugs.gnu.org, Leo Famulari <leo <at> famulari.name>
Subject: Re: bug#30467: GuixSD fails to display login prompt on headless server
Date: Fri, 16 Feb 2018 01:53:12 +0100

Hi George,

On Thu, 15 Feb 2018 16:29:11 -0500
George myglc2 Clemmer <myglc2 <at> gmail.com> wrote:

> Not clear what you wanted so I didn't try this.

Leo clarified that not specifying it is working, so it's fine not to specify it.

So please ignore.

> FWIW, I don't see agetty running ...

What happens when you run manually:

agetty /dev/ttyS1

Does it keep running?

Message #23 received at 30467 <at> debbugs.gnu.org (full text, mbox):

From: George myglc2 Clemmer <myglc2 <at> gmail.com>
To: Danny Milosavljevic <dannym <at> scratchpost.org>
Cc: 30467 <at> debbugs.gnu.org, Leo Famulari <leo <at> famulari.name>
Subject: Re: bug#30467: GuixSD fails to display login prompt on headless server
Date: Thu, 15 Feb 2018 20:38:38 -0500

Hi Danny,

On 02/16/2018 at 00:53 Danny Milosavljevic writes:

> Hi George,
>
> On Thu, 15 Feb 2018 16:29:11 -0500
> George myglc2 Clemmer <myglc2 <at> gmail.com> wrote:

>> FWIW, I don't see agetty running ...
>
> What happens when you run manually:
>
> agetty /dev/ttyS1
>
> Does it keep running?

No. /var/log/debug is also jammed like this ...

[...]
Feb 15 19:05:29 localhost vmunix: [    5.890367] Creating 1 MTD partitions on "intel-spi":
Feb 15 19:05:29 localhost vmunix: [    5.890370] 0x000000000000-0x000000800000 : "BIOS"
Feb 15 19:05:29 localhost vmunix: [    5.906526] [TTM] Initializing pool allocator
Feb 15 19:05:29 localhost vmunix: [    5.910911] [TTM] Initializing DMA pool allocator
Feb 15 19:05:29 localhost vmunix: [    6.023282] fbcon: astdrmfb (fb0) is primary device
Feb 15 19:05:29 localhost vmunix: [    6.091450] Console: switching to colour frame buffer device 160x64
Feb 15 19:05:29 localhost vmunix: [    6.099375] ipmi_si IPI0001:00: Found new BMC (man_id: 0x000000, prod_id: 0xaabb, dev_id: 0x20)
Feb 15 19:05:30 localhost vmunix: [    6.105342] pps pps0: new PPS source ptp0
Feb 15 19:05:30 localhost vmunix: [    6.105363] igb 0000:03:00.0: added PHC on eth0
Feb 15 19:05:30 localhost vmunix: [    6.105363] igb 0000:03:00.0: Intel(R) Gigabit Ethernet Network Connection
Feb 15 19:05:30 localhost vmunix: [    6.105365] igb 0000:03:00.0: eth0: (PCIe:2.5Gb/s:Width x1) d0:50:99:5b:d7:55
Feb 15 19:05:30 localhost vmunix: [    6.105959] igb 0000:03:00.0: eth0: PBA No: 001300-000
Feb 15 19:05:30 localhost vmunix: [    6.105972] igb 0000:03:00.0: Using MSI-X interrupts. 4 rx queue(s), 4 tx queue(s)
Feb 15 19:05:30 localhost vmunix: [    6.149879] ipmi_si IPI0001:00: IPMI kcs interface initialized
Feb 15 19:05:30 localhost vmunix: [    6.182426] ast 0000:06:00.0: fb0: astdrmfb frame buffer device
Feb 15 19:05:30 localhost vmunix: [    6.188672] IPMI SSIF Interface driver
Feb 15 19:05:30 localhost vmunix: [    6.205279] intel_rapl: Found RAPL domain package
Feb 15 19:05:30 localhost vmunix: [    6.209992] intel_rapl: Found RAPL domain core
Feb 15 19:05:30 localhost vmunix: [    6.211999] [drm] Initialized ast 0.1.0 20120228 for 0000:06:00.0 on minor 0
Feb 15 19:05:30 localhost vmunix: [    6.221483] intel_rapl: Found RAPL domain dram
Feb 15 19:05:30 localhost vmunix: [    6.225945] intel_rapl: RAPL package 0 domain package locked by BIOS
Feb 15 19:05:30 localhost vmunix: [    6.232290] intel_rapl: RAPL package 0 domain dram locked by BIOS
Feb 15 19:05:30 localhost vmunix: [    6.323457] Error: Driver 'pcspkr' is already registered, aborting...
Feb 15 19:05:30 localhost vmunix: [    6.351730] pps pps1: new PPS source ptp1
Feb 15 19:05:31 localhost vmunix: [    6.355748] igb 0000:04:00.0: added PHC on eth1
Feb 15 19:05:31 localhost vmunix: [    6.360278] igb 0000:04:00.0: Intel(R) Gigabit Ethernet Network Connection
Feb 15 19:05:31 localhost vmunix: [    6.367150] igb 0000:04:00.0: eth1: (PCIe:2.5Gb/s:Width x1) d0:50:99:5b:d7:54
Feb 15 19:05:31 localhost vmunix: [    6.374341] igb 0000:04:00.0: eth1: PBA No: 001300-000
Feb 15 19:05:31 localhost vmunix: [    6.379472] igb 0000:04:00.0: Using MSI-X interrupts. 4 rx queue(s), 4 tx queue(s)
Feb 15 19:05:31 localhost vmunix: [    6.393246] igb 0000:04:00.0 enp4s0: renamed from eth1
Feb 15 19:05:31 localhost vmunix: [    6.424048] igb 0000:03:00.0 enp3s0: renamed from eth0
Feb 15 19:05:31 localhost vmunix: [    6.457010] Adding 10117460k swap on /dev/nvme0n1p2.  Priority:-2 extents:1 across:10117460k SSFS
Feb 15 19:05:31 localhost vmunix: [    7.559634] IPv6: ADDRCONF(NETDEV_UP): enp3s0: link is not ready
Feb 15 19:05:31 localhost vmunix: [   10.412657] igb 0000:03:00.0 enp3s0: igb: enp3s0 NIC Link is Up 1000 Mbps Full Duplex, Flow Control: RX/TX
Feb 15 19:05:31 localhost vmunix: [   10.528283] IPv6: ADDRCONF(NETDEV_CHANGE): enp3s0: link becomes ready
Feb 15 19:08:06 localhost ntpd[368]: receive: Unexpected origin timestamp 0xde309f65.f5541f08 does not match aorg 0000000000.00000000 from server <at> 162.210.110.4 xmt 0xde309f66.6426f9e7
Feb 15 19:10:26 localhost vmunix: [  349.762770] sysrq: SysRq : HELP : loglevel(0-9) reboot(b) crash(c) terminate-all-tasks(e) memory-full-oom-kill(f) kill-all-tasks(i) thaw-filesystems(j) sak(k) show-backtrace-all-active-cpus(l) show-memory-usage(m) nice-all-RT-tasks(n) poweroff(o) show-registers(p) show-all-timers(q) unraw(r) sync(s) show-task-states(t) unmount(u) force-fb(V) show-blocked-tasks(w) dump-ftrace-buffer(z) 
Feb 15 19:13:39 localhost ntpd[368]: kernel reports TIME_ERROR: 0x41: Clock Unsynchronized
Feb 15 19:24:44 localhost -- MARK --
Feb 15 19:44:44 localhost -- MARK --
Feb 15 19:59:03 localhost vmunix: [ 3266.992550] sysrq: SysRq : HELP : loglevel(0-9) reboot(b) crash(c) terminate-all-tasks(e) memory-full-oom-kill(f) kill-all-tasks(i) thaw-filesystems(j) sak(k) show-backtrace-all-active-cpus(l) show-memory-usage(m) nice-all-RT-tasks(n) poweroff(o) show-registers(p) show-all-timers(q) unraw(r) sync(s) show-task-states(t) unmount(u) force-fb(V) show-blocked-tasks(w) dump-ftrace-buffer(z) 
Feb 15 20:13:32 localhost vmunix: [ 4135.206111] sysrq: SysRq : HELP : loglevel(0-9) reboot(b) crash(c) terminate-all-tasks(e) memory-full-oom-kill(f) kill-all-tasks(i) thaw-filesystems(j) sak(k) show-backtrace-all-active-cpus(l) show-memory-usage(m) nice-all-RT-tasks(n) poweroff(o) show-registers(p) show-all-timers(q) unraw(r) sync(s) show-task-states(t) unmount(u) force-fb(V) show-blocked-tasks(w) dump-ftrace-buffer(z) 
Feb 15 20:24:44 localhost -- MARK --

Message #26 received at 30467 <at> debbugs.gnu.org (full text, mbox):

From: Danny Milosavljevic <dannym <at> scratchpost.org>
To: George myglc2 Clemmer <myglc2 <at> gmail.com>
Cc: 30467 <at> debbugs.gnu.org, Leo Famulari <leo <at> famulari.name>
Subject: Re: bug#30467: GuixSD fails to display login prompt on headless server
Date: Sat, 17 Feb 2018 18:04:09 +0100

Huh, please try

strace -f agetty /dev/ttyS1 >log 2>&1

and send the log to the list...

Message #29 received at 30467 <at> debbugs.gnu.org (full text, mbox):

From: Danny Milosavljevic <dannym <at> scratchpost.org>
To: George myglc2 Clemmer <myglc2 <at> gmail.com>
Cc: 30467 <at> debbugs.gnu.org, Leo Famulari <leo <at> famulari.name>
Subject: Re: bug#30467: GuixSD fails to display login prompt on headless server
Date: Sat, 17 Feb 2018 18:05:30 +0100

What does

stty -F /dev/ttyS1

say?

Message #32 received at 30467 <at> debbugs.gnu.org (full text, mbox):

From: George myglc2 Clemmer <myglc2 <at> gmail.com>
To: Danny Milosavljevic <dannym <at> scratchpost.org>
Cc: 30467 <at> debbugs.gnu.org, Leo Famulari <leo <at> famulari.name>
Subject: Re: bug#30467: GuixSD fails to display login prompt on headless server
Date: Wed, 21 Feb 2018 20:35:34 -0500

[Message part 1 (text/plain, inline)]

Hi Danny,

I think we got side-tracked to the serial port when my real issue is
with the VGA port. Maybe my bug report was not clear. Let me clarify ...

My headless server stopped displaying the console tty login *on the VGA
port* somewhere between (GNU Guix) 0.13.0.4589-74bea6 (my G18) and (GNU
Guix) 0.13.0.5042-6e385 (My G19). In the later version, when I reboot
the VGA displays 1) the BIOS prompt, 2) the grub boot menu, 3) "Booting
GNU with Linuxlibre ..." for a couple seconds, and then 4) a blank.

In either version, this config ...

(agetty-service (agetty-configuration (tty "ttyS1")
                (baud-rate "115200")))

... causes a login prompt to be raised on ttyS1 which, on my system, is
mapped to IPMI Serial over Lan (SOL). However recently you said ...

On 02/15/2018 at 15:54 Danny Milosavljevic writes:

> In the long run please remove your agetty-service, it should automatically appear
> (it recovers the settings from the "console" kernel-argument).

So I tried that with a build from today (guix (GNU Guix)
0.13.0.5116-e272), and *no login* is raised on ttyS1 (My G46, sys.scm
attached).

However, if I *ignore your advice* a login prompt *is raised* on ttyS1
(My G45).

So, IIUC, it's not working the way you said it would, but it's not a
problem for me because I can just use the agetty-service.

Not sure your questions are still relevant, but here is what I see ...

On 02/17/2018 at 17:05 Danny Milosavljevic writes:

> What does
>
> stty -F /dev/ttyS1
>
> say?

g1 <at> g1 ~/www$ sudo stty -F /dev/ttyS1
speed 115200 baud; line = 0;
-brkint -imaxbel

On 02/17/2018 at 17:04 Danny Milosavljevic writes:

> Huh, please try
>
> strace -f agetty /dev/ttyS1 >log 2>&1
>
> and send the log to the list...

(log.gz attached)

HTH - George

[log.gz (application/octet-stream, attachment)]

[sys.scm (application/octet-stream, attachment)]

Message #35 received at 30467 <at> debbugs.gnu.org (full text, mbox):

From: George myglc2 Clemmer <myglc2 <at> gmail.com>
To: Danny Milosavljevic <dannym <at> scratchpost.org>
Cc: 30467 <at> debbugs.gnu.org, Leo Famulari <leo <at> famulari.name>
Subject: Re: bug#30467: GuixSD fails to display login prompt on headless server
Date: Wed, 21 Feb 2018 22:29:40 -0500

In my previous message, please disregard comments that begin with ...

> So I tried that with a build from today (guix (GNU Guix)
> 0.13.0.5116-e272), and *no login* is raised on ttyS1 (My G46, sys.scm
> attached).

which is mistakenly based on an old commit :-|

Message #38 received at 30467 <at> debbugs.gnu.org (full text, mbox):

From: Danny Milosavljevic <dannym <at> scratchpost.org>
To: George myglc2 Clemmer <myglc2 <at> gmail.com>
Cc: 30467 <at> debbugs.gnu.org, Leo Famulari <leo <at> famulari.name>
Subject: Re: bug#30467: GuixSD fails to display login prompt on headless server
Date: Thu, 22 Feb 2018 12:41:39 +0100

Hi George,

On Wed, 21 Feb 2018 20:35:34 -0500
George myglc2 Clemmer <myglc2 <at> gmail.com> wrote:

> I think we got side-tracked to the serial port when my real issue is
> with the VGA port. Maybe my bug report was not clear. Let me clarify ...

Oh, that changes things.  Ludo is debugging some silly UTF-8 problem
with the virtual terminals on the VGA port right now - might be related,
or not.  Who knows...

> GNU with Linuxlibre ..." for a couple seconds, and then 4) a blank.

Is there still a cursor?

What happens when you don't use the serial (so no console=...) and then
try to write something to /dev/console ?  It should show up somewhere.
Where?  On the VGA port?

Let's try to find where in the chain it goes off the rails.

If writing to /dev/console manually works, try

ps -ef | grep mingetty

Are they running?

Also,

ls -l /dev/tty[123456] /dev/tty

> In either version, this config ...
> 
> (agetty-service (agetty-configuration (tty "ttyS1")
>                 (baud-rate "115200")))
> ... causes a login prompt to be raised on ttyS1 which, on my system, is
> mapped to IPMI Serial over Lan (SOL).

Yeah, okay.  So the actual tty-login-in works... *phiew*.

> g1 <at> g1 ~/www$ sudo stty -F /dev/ttyS1
> speed 115200 baud; line = 0;
> -brkint -imaxbel

Looks good - login also works fine on the serial, so all is well there... :)

Now let's see what the problem with the VGA is...

Message #41 received at 30467 <at> debbugs.gnu.org (full text, mbox):

From: George myglc2 Clemmer <myglc2 <at> gmail.com>
To: Danny Milosavljevic <dannym <at> scratchpost.org>
Cc: 30467 <at> debbugs.gnu.org, Leo Famulari <leo <at> famulari.name>
Subject: Re: bug#30467: GuixSD fails to display login prompt on headless server
Date: Fri, 23 Feb 2018 10:46:42 -0500

On 02/22/2018 at 11:41 Danny Milosavljevic writes:

> Hi George,
>
> On Wed, 21 Feb 2018 20:35:34 -0500
> George myglc2 Clemmer <myglc2 <at> gmail.com> wrote:
>
>> I think we got side-tracked to the serial port when my real issue is
>> with the VGA port. Maybe my bug report was not clear. Let me clarify ...
>
> Oh, that changes things.  Ludo is debugging some silly UTF-8 problem
> with the virtual terminals on the VGA port right now - might be related,
> or not.  Who knows...

Well this problem goes back to November. The problem first appears in
my generation 19 ...

     19            GNU with Linux-Libre 4.14.2 (beta)           61 2017-11-30 19:18:15
     18            GNU with Linux-Libre 4.13.12 (beta)          62 2017-11-11 13:05:56

... which  correspond to ...

19 (GNU Guix) 0.13.0.5042-6e385
18 (GNU Guix) 0.13.0.4589-74bea6

Note: The results cited below are WRT my most recent build done
yesterday, guix (GNU Guix) 0.14.0.2455-26737.

>> GNU with Linuxlibre ..." for a couple seconds, and then 4) a blank.
>
> Is there still a cursor?

NO.

> What happens when you don't use the serial (so no console=...)

The VGA displays ...

- BIOS prompt
- grub menu
- 'Booting GNU linux-libre ...'
- ~1 seconds of messages
- Blank screen w/ no cursor

... which differers in that there are a couple seconds of boot messages.

> and then
> try to write something to /dev/console ?  It should show up somewhere.
> Where?  On the VGA port?

/home/g1# echo 123 > /dev/console

... shows nothing on the VGA or serial.

Message #44 received at 30467 <at> debbugs.gnu.org (full text, mbox):

From: Danny Milosavljevic <dannym <at> scratchpost.org>
To: George myglc2 Clemmer <myglc2 <at> gmail.com>
Cc: 30467 <at> debbugs.gnu.org
Subject: Re: bug#30467: GuixSD fails to display login prompt on headless server
Date: Sat, 24 Feb 2018 15:51:34 +0100

[Message part 1 (text/plain, inline)]

> > Is there still a cursor?  
> 
> NO.

That's... bad.

The tool I'd need to debug this (fbset) is not available in Guix.

And my guix git checkout is a little under construction.

So can you please try to guix package -f the attached file?

Then please run fbset.

Does it work?  What does it say?

[fbset.scm (text/x-scheme, attachment)]

Message #47 received at 30467 <at> debbugs.gnu.org (full text, mbox):

From: George myglc2 Clemmer <myglc2 <at> gmail.com>
To: Danny Milosavljevic <dannym <at> scratchpost.org>
Cc: 30467 <at> debbugs.gnu.org
Subject: Re: bug#30467: GuixSD fails to display login prompt on headless server
Date: Mon, 26 Feb 2018 12:02:11 -0500

Hi Danny,

On 02/24/2018 at 14:51 Danny Milosavljevic writes:

>> > Is there still a cursor?  
>> 
>> NO.
>
> That's... bad.
>
> The tool I'd need to debug this (fbset) is not available in Guix.
>
> And my guix git checkout is a little under construction.
>
> So can you please try to guix package -f the attached file?

Tried to but got this ...

g1 <at> g1 ~/bug/bug#30467-vga-no-login$ guix package -f fbset.scm 
The following package will be installed:
   fbset	2.1	/gnu/store/kg15k6llj062gwg3jzymimas93pwwysw-fbset-2.1
[..]
phase `patch-usr-bin-file' succeeded after 0.0 seconds
starting phase `patch-source-shebangs'
patch-shebang: ./modeline2fb: warning: no binary for interpreter `perl' found in $PATH
phase `patch-source-shebangs' succeeded after 0.0 seconds
starting phase `patch-generated-file-shebangs'
patch-shebang: ./modeline2fb: warning: no binary for interpreter `perl' found in $PATH
phase `patch-generated-file-shebangs' succeeded after 0.0 seconds
starting phase `build'
gcc -Wall -O2 -I.    -c -o fbset.o fbset.c
bison -d modes.y
flex modes.l
make: *** No rule to make target 'modes.tab.h', needed by 'lex.yy.o'.  Stop.
make: *** Waiting for unfinished jobs....
phase `build' failed after 0.2 seconds
builder for `/gnu/store/hf71jn7x88dm77x1368ajai4l0gp276c-fbset-2.1.drv' failed with exit code 1
cannot build derivation `/gnu/store/nigbs3x4sxpix1dz16r1ypvkmp7s7p2c-profile.drv': 1 dependencies couldn't be built
guix package: error: build failed: build of `/gnu/store/nigbs3x4sxpix1dz16r1ypvkmp7s7p2c-profile.drv' failed

g1 <at> g1 ~/bug/bug#30467-vga-no-login$ which perl
/home/g1/.guix-profile/bin/perl

>
> Then please run fbset.
>
> Does it work?  What does it say?

TIA - George

Message #50 received at 30467 <at> debbugs.gnu.org (full text, mbox):

From: Danny Milosavljevic <dannym <at> scratchpost.org>
To: George myglc2 Clemmer <myglc2 <at> gmail.com>
Cc: 30467 <at> debbugs.gnu.org
Subject: Re: bug#30467: GuixSD fails to display login prompt on headless server
Date: Mon, 26 Feb 2018 18:32:00 +0100

[Message part 1 (text/plain, inline)]

Try the attached one...

[fbset.scm (text/x-scheme, attachment)]

Message #53 received at 30467 <at> debbugs.gnu.org (full text, mbox):

From: George myglc2 Clemmer <myglc2 <at> gmail.com>
To: Danny Milosavljevic <dannym <at> scratchpost.org>
Cc: 30467 <at> debbugs.gnu.org
Subject: Re: bug#30467: GuixSD fails to display login prompt on headless server
Date: Mon, 26 Feb 2018 13:50:35 -0500

On 02/26/2018 at 17:32 Danny Milosavljevic writes:

> Try the attached one...

g1 <at> g1 ~/bug/bug#30467-vga-no-login$ sudo fbset

mode "1280x1024"
    geometry 1280 1024 1280 1024 32
    timings 0 0 0 0 0 0 0
    accel true
    rgba 8/16,8/8,8/0,0/0
endmode

g1 <at> g1 ~/bug/bug#30467-vga-no-login$ 

Message #56 received at 30467 <at> debbugs.gnu.org (full text, mbox):

From: Danny Milosavljevic <dannym <at> scratchpost.org>
To: George myglc2 Clemmer <myglc2 <at> gmail.com>
Cc: 30467 <at> debbugs.gnu.org
Subject: Re: bug#30467: GuixSD fails to display login prompt on headless server
Date: Tue, 27 Feb 2018 13:08:56 +0100

> g1 <at> g1 ~/bug/bug#30467-vga-no-login$ sudo fbset
> 
> mode "1280x1024"
>     geometry 1280 1024 1280 1024 32
>     timings 0 0 0 0 0 0 0
>     accel true
>     rgba 8/16,8/8,8/0,0/0
> endmode
> 
> g1 <at> g1 ~/bug/bug#30467-vga-no-login$ 

Ok, so there is a screen there...

Does

sudo -i
echo foo >/dev/tty1

work and print something there?

What is output of the command "dmesg" ?

Also, please try

ps -ef |grep mingetty

Message #59 received at 30467 <at> debbugs.gnu.org (full text, mbox):

From: George myglc2 Clemmer <myglc2 <at> gmail.com>
To: Danny Milosavljevic <dannym <at> scratchpost.org>
Cc: 30467 <at> debbugs.gnu.org
Subject: Re: bug#30467: GuixSD fails to display login prompt on headless server
Date: Tue, 27 Feb 2018 08:39:08 -0500

[Message part 1 (text/plain, inline)]

On 02/27/2018 at 12:08 Danny Milosavljevic writes:

> Does
>
> sudo -i
> echo foo >/dev/tty1
>
> work and print something there?

Shows no error and nothing on the VGA.

> What is output of the command "dmesg" ?
>
> Also, please try
>
> ps -ef |grep mingetty

Please see attached.

TIA, George

[dmesg.log.gz (application/octet-stream, attachment)]

Message #62 received at 30467 <at> debbugs.gnu.org (full text, mbox):

From: myglc2 <at> gmail.com
To: 30467 <at> debbugs.gnu.org
Subject: Re: bug#30467: GuixSD fails to display login prompt on headless server
Date: Thu, 08 Mar 2018 09:16:06 -0500

commit ...
c32e3dded * services: agetty: Call default-serial-port only when starting.
... fixed this here, thanks - George

Message #67 received at 30467-done <at> debbugs.gnu.org (full text, mbox):

From: ludo <at> gnu.org (Ludovic Courtès)
To: myglc2 <at> gmail.com
Cc: Danny Milosavljevic <dannym <at> scratchpost.org>, 30467-done <at> debbugs.gnu.org
Subject: Re: bug#30467: GuixSD fails to display login prompt on headless server
Date: Sat, 10 Mar 2018 00:23:34 +0100

myglc2 <at> gmail.com skribis:

> commit ...
> c32e3dded * services: agetty: Call default-serial-port only when starting.
> ... fixed this here, thanks - George

Awesome, thanks for confirming.

And thanks Danny for the fix!

Ludo’.

Previous Next

GNU bug tracking system
Copyright (C) 1999 Darren O. Benham, 1997,2003 nCipher Corporation Ltd, 1994-97 Ian Jackson.