[fedora-arm] Re: F25 beta failure with RPi-3

[Dexter Haslem]

I've been chasing this same issue and reported it (albeit lacking details)
on previous LXDE spin beta [1]. I can reproduce on latest image,
using same kernel as Richard, however sshd doesn't appear to be up
for me so I'm having trouble getting a dmesg off the device. If I have
any luck I'll put it up here. I was also using HDMI out, to a ASUS VX238
(1920x1080 native)


Dexter

[1] https://bugzilla.redhat.com/show_bug.cgi?id=1384295
___
arm mailing list -- arm@lists.fedoraproject.org
To unsubscribe send an email to arm-le...@lists.fedoraproject.org

[fedora-arm] Re: F25 beta failure with RPi-3

[Richard Ryniker]

Two more trials without any improvement:

  Using kernel-4.9.0-0.rc2.git0.2.fc26 does not help.  I had no reason to
  think it would, but now that ssh to my RPi3 works well, it was easy to
  try.

  Changing the display to a Dell P2415Q (3840x2160 resolution, HDMI
  input) does not help.  Again, no reason to think it would, but the
  world has a lot of quirky hardware and this helps me believe it is not
  a display device issue.

The last line I see before the display goes blank is:

  fb: switching to vc4drmfb from simple

which is no surprise.

Journal and dmesg files are at:  http://ryniker.org/Fedora/arm/
___
arm mailing list -- arm@lists.fedoraproject.org
To unsubscribe send an email to arm-le...@lists.fedoraproject.org

[fedora-arm] Re: F25 beta failure with RPi-3

[Richard Ryniker]

Peter Robinson  on Tue, 25 Oct 2016 10:23:13 wrote:

>There's a new 4.8.4-301.fc25 kernel that I think should fix this issue

No improvement with this new kernel.  I set default to multi-user.target
in an attempt to avoid any X11 confusion.  After several screens of boot
messages scroll past, the display goes dark (no loss of input signal -
the monitor would report that.)

Display device is Viewsonic model VX2235wm (native resolution 1680x1050) with
DVI-D input connected to my RPi HDMI port.

Is this kernel supposed to avoid the "i2c-bcm2835 3f805000.i2c: i2c
transfer failed: 100" messages that clutter dmesg and the journal?

I have not tried to perform any direct measurement, but the time when the
display goes dark seems to be near 27 seconds when these messages first
appear.

dmesg output:

[0.00] Booting Linux on physical CPU 0x0
[0.00] Linux version 4.8.4-301.fc25.armv7hl 
(mockbu...@arm04-builder06.arm.fedoraproject.org) (gcc version 6.2.1 20160916 
(Red Hat 6.2.1-2) (GCC) ) #1 SMP Tue Oct 25 02:01:39 UTC 2016
[0.00] CPU: ARMv7 Processor [410fd034] revision 4 (ARMv7), cr=10c5383d
[0.00] CPU: div instructions available: patching division code
[0.00] CPU: PIPT / VIPT nonaliasing data cache, VIPT aliasing 
instruction cache
[0.00] OF: fdt:Machine model: Raspberry Pi 3 Model B
[0.00] efi: Getting EFI parameters from FDT:
[0.00] efi: UEFI not found.
[0.00] cma: Reserved 256 MiB at 0x2000
[0.00] Memory policy: Data cache writealloc
[0.00] On node 0 totalpages: 253952
[0.00] free_area_init_node: node 0, pgdat c0e66b80, node_mem_map 
df6c1000
[0.00]   Normal zone: 1728 pages used for memmap
[0.00]   Normal zone: 0 pages reserved
[0.00]   Normal zone: 196608 pages, LIFO batch:31
[0.00]   HighMem zone: 57344 pages, LIFO batch:15
[0.00] percpu: Embedded 15 pages/cpu @df67a000 s31500 r8192 d21748 
u61440
[0.00] pcpu-alloc: s31500 r8192 d21748 u61440 alloc=15*4096
[0.00] pcpu-alloc: [0] 0 [0] 1 [0] 2 [0] 3 
[0.00] Built 1 zonelists in Zone order, mobility grouping on.  Total 
pages: 252224
[0.00] Kernel command line: ro 
root=UUID=6ce3f499-c222-4d7b-a3d6-a0a4e28340a5  cma=192MB cma=256MB 
LANG=en_US.UTF-8
[0.00] PID hash table entries: 4096 (order: 2, 16384 bytes)
[0.00] Dentry cache hash table entries: 131072 (order: 7, 524288 bytes)
[0.00] Inode-cache hash table entries: 65536 (order: 6, 262144 bytes)
[0.00] Memory: 715032K/1015808K available (7032K kernel code, 1092K 
rwdata, 3288K rodata, 1160K init, 853K bss, 38632K reserved, 262144K 
cma-reserved, 229376K highmem)
[0.00] Virtual kernel memory layout:
   vector  : 0x - 0x1000   (   4 kB)
   fixmap  : 0xffc0 - 0xfff0   (3072 kB)
   vmalloc : 0xf080 - 0xff80   ( 240 MB)
   lowmem  : 0xc000 - 0xf000   ( 768 MB)
   pkmap   : 0xbfe0 - 0xc000   (   2 MB)
   modules : 0xbf00 - 0xbfe0   (  14 MB)
 .text : 0xc0208000 - 0xc08e6534   (7034 kB)
 .init : 0xc0c56000 - 0xc0d78000   (1160 kB)
 .data : 0xc0d78000 - 0xc0e89184   (1093 kB)
  .bss : 0xc0e89184 - 0xc0f5e730   ( 854 kB)
[0.00] SLUB: HWalign=64, Order=0-3, MinObjects=0, CPUs=4, Nodes=1
[0.00] Hierarchical RCU implementation.
[0.00]  Build-time adjustment of leaf fanout to 32.
[0.00]  RCU restricting CPUs from NR_CPUS=32 to nr_cpu_ids=4.
[0.00] RCU: Adjusting geometry for rcu_fanout_leaf=32, nr_cpu_ids=4
[0.00] NR_IRQS:16 nr_irqs:16 16
[0.16] sched_clock: 32 bits at 1000kHz, resolution 1000ns, wraps every 
2147483647500ns
[0.40] clocksource: timer: mask: 0x max_cycles: 0x, 
max_idle_ns: 1911260446275 ns
[0.000122] bcm2835: system timer (irq = 33)
[0.001390] arm_arch_timer: WARNING: Invalid trigger for IRQ16, assuming 
level low
[0.001400] arm_arch_timer: WARNING: Please fix your firmware
[0.001415] arm_arch_timer: WARNING: Invalid trigger for IRQ17, assuming 
level low
[0.001424] arm_arch_timer: WARNING: Please fix your firmware
[0.001555] arm_arch_timer: Architected cp15 timer(s) running at 19.20MHz 
(phys).
[0.001575] clocksource: arch_sys_counter: mask: 0xff 
max_cycles: 0x46d987e47, max_idle_ns: 440795202767 ns
[0.001597] sched_clock: 56 bits at 19MHz, resolution 52ns, wraps every 
4398046511078ns
[0.001614] Switching to timer-based delay loop, resolution 52ns
[0.002080] Console: colour dummy device 80x30
[0.003737] console [tty0] enabled
[0.003827] Calibrating delay loop (skipped), value calculated using timer 
frequency.. 38.40 BogoMIPS (lpj=96000)
[0.003914] pid_max: default: 32768 minimum: 301
[0.004440] 

[fedora-arm] Re: F25 beta failure with RPi-3

[Peter Robinson]

On Mon, Oct 24, 2016 at 9:00 PM, Richard Ryniker  wrote:
> Peter Robinson  wrote on Sat, 22 Oct 2016 15:18:00:
>
>>Can you just provide  the output of dmesg?
>
> Yes, below, but it was a chore.  How does one execute dmesg when there is
> no accessible user?  The system boots far enough to initialize the
> Ethernet connection, and sshd is active, but all of the defined users
> have no (or unknown) passwords.  There are very good reasons for this, I
> agree with them, it just makes strange cases like mine more difficult.
>
> Failed solution 1:  edit /etc/passwd to allow login without a password.
> Either sshd or login is unhappy and forbids login.

Updating the /etc/shadow line for root to add a known password hash is
all I usually do and it has worked in the past.

There's a new 4.8.4-301.fc25 kernel that I think should fix this issue
so you can grab that from koji [1] or it'll be going out to
updates-testing probably today and should be stable by the end of the
week.

Also for reference what is the monitor, the type of connector used and
the resolution it runs at?

Peter

[1] http://koji.fedoraproject.org/koji/buildinfo?buildID=812331
___
arm mailing list -- arm@lists.fedoraproject.org
To unsubscribe send an email to arm-le...@lists.fedoraproject.org

[fedora-arm] Re: F25 beta failure with RPi-3

[Richard Ryniker]

Peter Robinson  wrote on Sat, 22 Oct 2016 15:18:00:

>Can you just provide  the output of dmesg?

Yes, below, but it was a chore.  How does one execute dmesg when there is
no accessible user?  The system boots far enough to initialize the
Ethernet connection, and sshd is active, but all of the defined users
have no (or unknown) passwords.  There are very good reasons for this, I
agree with them, it just makes strange cases like mine more difficult.

Failed solution 1:  edit /etc/passwd to allow login without a password.
Either sshd or login is unhappy and forbids login.

Failed solution 2:  use pwunconv on another machine.  Did not work (I do
not know why; probably some configuration parameter.)  pwunconv did not
complain; it simply did nothing.

Working solution 3: manually add a new user to /etc/passwd and
/etc/shadow with a real password.  Add this user to /etc/sudoers.  Now
sshd is happy, and remote login can acquire root privilege.

With this capability,  I performed a dnf upgrade (On October 22) to
install the latest X server and the newest kernel.

The original problem continues: post-installation user setup (gdm?) fails
- dark screen.  This is an intriguing problem, because I would expect the
Raspberry Pi to be a far less variable environment that other machines
that have many different BIOS and hardware configurations.  For RPi 3,
just plug in an HDMI cable and go...

My RPi board is functional: it boots Raspbian without problem.

I do not see any smoking gun in the dmesg output, but I do not have an
example of F25 working that might direct attention to something different
in this case.


-bash-4.3$ dmesg
[0.00] Booting Linux on physical CPU 0x0
[0.00] Linux version 4.8.3-300.fc25.armv7hl 
(mockbu...@arm04-builder05.arm.fedoraproject.org) (gcc version 6.2.1 20160916 
(Red Hat 6.2.1-2) (GCC) ) #1 SMP Thu Oct 20 16:49:07 UTC 2016
[0.00] CPU: ARMv7 Processor [410fd034] revision 4 (ARMv7), cr=10c5383d
[0.00] CPU: div instructions available: patching division code
[0.00] CPU: PIPT / VIPT nonaliasing data cache, VIPT aliasing 
instruction cache
[0.00] OF: fdt:Machine model: Raspberry Pi 3 Model B
[0.00] efi: Getting EFI parameters from FDT:
[0.00] efi: UEFI not found.
[0.00] cma: Reserved 256 MiB at 0x2000
[0.00] Memory policy: Data cache writealloc
[0.00] On node 0 totalpages: 253952
[0.00] free_area_init_node: node 0, pgdat c0e66b00, node_mem_map 
df6c1000
[0.00]   Normal zone: 1728 pages used for memmap
[0.00]   Normal zone: 0 pages reserved
[0.00]   Normal zone: 196608 pages, LIFO batch:31
[0.00]   HighMem zone: 57344 pages, LIFO batch:15
[0.00] percpu: Embedded 15 pages/cpu @df67a000 s31500 r8192 d21748 
u61440
[0.00] pcpu-alloc: s31500 r8192 d21748 u61440 alloc=15*4096
[0.00] pcpu-alloc: [0] 0 [0] 1 [0] 2 [0] 3 
[0.00] Built 1 zonelists in Zone order, mobility grouping on.  Total 
pages: 252224
[0.00] Kernel command line: ro 
root=UUID=6ce3f499-c222-4d7b-a3d6-a0a4e28340a5  cma=192MB cma=256MB 
LANG=en_US.UTF-8
[0.00] PID hash table entries: 4096 (order: 2, 16384 bytes)
[0.00] Dentry cache hash table entries: 131072 (order: 7, 524288 bytes)
[0.00] Inode-cache hash table entries: 65536 (order: 6, 262144 bytes)
[0.00] Memory: 715000K/1015808K available (7032K kernel code, 1092K 
rwdata, 3288K rodata, 1160K init, 853K bss, 38664K reserved, 262144K 
cma-reserved, 229376K highmem)
[0.00] Virtual kernel memory layout:
   vector  : 0x - 0x1000   (   4 kB)
   fixmap  : 0xffc0 - 0xfff0   (3072 kB)
   vmalloc : 0xf080 - 0xff80   ( 240 MB)
   lowmem  : 0xc000 - 0xf000   ( 768 MB)
   pkmap   : 0xbfe0 - 0xc000   (   2 MB)
   modules : 0xbf00 - 0xbfe0   (  14 MB)
 .text : 0xc0208000 - 0xc08e62b4   (7033 kB)
 .init : 0xc0c56000 - 0xc0d78000   (1160 kB)
 .data : 0xc0d78000 - 0xc0e89104   (1093 kB)
  .bss : 0xc0e89104 - 0xc0f5e6b0   ( 854 kB)
[0.00] SLUB: HWalign=64, Order=0-3, MinObjects=0, CPUs=4, Nodes=1
[0.00] Hierarchical RCU implementation.
[0.00]  Build-time adjustment of leaf fanout to 32.
[0.00]  RCU restricting CPUs from NR_CPUS=32 to nr_cpu_ids=4.
[0.00] RCU: Adjusting geometry for rcu_fanout_leaf=32, nr_cpu_ids=4
[0.00] NR_IRQS:16 nr_irqs:16 16
[0.17] sched_clock: 32 bits at 1000kHz, resolution 1000ns, wraps every 
2147483647500ns
[0.44] clocksource: timer: mask: 0x max_cycles: 0x, 
max_idle_ns: 1911260446275 ns
[0.000122] bcm2835: system timer (irq = 33)
[0.001393] arm_arch_timer: WARNING: Invalid trigger for IRQ16, assuming 
level low
[0.001403] 

[fedora-arm] Re: F25 beta failure with RPi-3

[Peter Robinson]

On Thu, Oct 20, 2016 at 8:01 PM, Richard Ryniker  wrote:
> Boots (a few screens of messages) but fails to run first boot user setup.
>
> There appears to be some confusion about the output display (HDMI
> connection) though it worked through u-boot and beyond the switch to
> framebuffer for additional line output.
>
> Journalctl output (from the 2016.10.18 nightly) is available at:
>
>   http://ryniker.org/Fedora/arm/
>
> with the tail of the journal (1315 lines) at:
>
>   http://ryniker.org/Fedora/arm/pi-2016.10.18-journal_4-end

Can you just provide  the output of dmesg?

> Most obvious suspicious messages:
>
>   Oct 20 12:27:01 localhost.localdomain gnome-shell[1068]: JS ERROR: 
> TypeError: this.primaryMonitor is undefined
>
> LayoutManager<._updateBoxes@resource:///org/gnome/shell/ui/layout.js:461
>
> wrapper@resource:///org/gnome/gjs/modules/lang.js:178
>
> LayoutManager<._monitorsChanged@resource:///org/gnome/shell/ui/layout.js:495
>
> wrapper@resource:///org/gnome/gjs/modules/lang.js:178
>
> LayoutManager<._init@resource:///org/gnome/shell/ui/layout.js:277
>
> wrapper@resource:///org/gnome/gjs/modules/lang.js:178
>
> _Base.prototype._construct@resource:///org/gnome/gjs/modules/lang.js:110
>
> Class.prototype._construct/newClass@resource:///org/gnome/gjs/modules/lang.js:213
>
> _initializeUI@resource:///org/gnome/shell/ui/main.js:151
>
> start@resource:///org/gnome/shell/ui/main.js:125
>@:1
>   Oct 20 12:27:01 localhost.localdomain gnome-shell[1068]: Execution of 
> main.js threw exception: JS_EvaluateScript() failed
>   Oct 20 12:27:01 localhost.localdomain gnome-session[999]: 
> gnome-session-binary[999]: WARNING: App 'setup-shell.desktop' exited with 
> code 1
>   Oct 20 12:27:01 localhost.localdomain gnome-session-binary[999]: WARNING: 
> App 'setup-shell.desktop' exited with code 1
>
> but I also wonder about multiple earlier messages that say:
>
>   Oct 20 12:26:56 localhost.localdomain kernel: i2c-bcm2835 3f805000.i2c: i2c 
> transfer failed: 100

Just spurious messages and can be ignored.
___
arm mailing list -- arm@lists.fedoraproject.org
To unsubscribe send an email to arm-le...@lists.fedoraproject.org