Bug#792746: Upgrading to 4.0.8-1 kernel makes external monitors unusable

[Nathaniel Smith]

Yes, I haven't seen one of these crashes in quite some time -- sorry for
not updating.
On Jul 12, 2016 6:54 PM, "Ben Hutchings"  wrote:

> On Mon, 2016-07-11 at 22:45 +0200, Enrico Zini wrote:
> > On Sat, May 28, 2016 at 06:21:34PM +0200, Enrico Zini wrote:
> >
> > > When the problem happens, this appears on dmesg:
> > > [227725.132861] [drm:ironlake_irq_handler [i915]] *ERROR* PCH
> > > transcoder B FIFO underrun
> >
> > This seems fixed in 4.6.0-1-amd64:
> >
> >$ uptime
> > 22:44:00 up 6 days,  7:52,  1 user,  load average: 0.28, 0.34, 0.24
> >$ dmesg|grep transcoder
> >$
>
> Is this problem also fixed on your system, Nathaniel?
>
> Ben.
>
> --
>
> Ben Hutchings
> Sturgeon's Law: Ninety percent of everything is crap.
>

Bug#792746: Upgrading to 4.0.8-1 kernel makes external monitors unusable

[Ben Hutchings]

On Mon, 2016-07-11 at 22:45 +0200, Enrico Zini wrote:
> On Sat, May 28, 2016 at 06:21:34PM +0200, Enrico Zini wrote:
> 
> > When the problem happens, this appears on dmesg:
> > [227725.132861] [drm:ironlake_irq_handler [i915]] *ERROR* PCH
> > transcoder B FIFO underrun
> 
> This seems fixed in 4.6.0-1-amd64:
> 
>    $ uptime
> 22:44:00 up 6 days,  7:52,  1 user,  load average: 0.28, 0.34, 0.24
>    $ dmesg|grep transcoder
>    $ 

Is this problem also fixed on your system, Nathaniel?

Ben.

-- 

Ben Hutchings
Sturgeon's Law: Ninety percent of everything is crap.


signature.asc
Description: This is a digitally signed message part

Bug#792746: Upgrading to 4.0.8-1 kernel makes external monitors unusable

[Enrico Zini]

On Sat, May 28, 2016 at 06:21:34PM +0200, Enrico Zini wrote:

> When the problem happens, this appears on dmesg:
> [227725.132861] [drm:ironlake_irq_handler [i915]] *ERROR* PCH transcoder 
> B FIFO underrun

This seems fixed in 4.6.0-1-amd64:

   $ uptime
22:44:00 up 6 days,  7:52,  1 user,  load average: 0.28, 0.34, 0.24
   $ dmesg|grep transcoder
   $ 


Enrico

-- 
GPG key: 4096R/634F4BD1E7AD5568 2009-05-08 Enrico Zini 


signature.asc
Description: PGP signature

Bug#792746: Upgrading to 4.0.8-1 kernel makes external monitors unusable

[Enrico Zini]

On Sat, May 28, 2016 at 06:21:34PM +0200, Enrico Zini wrote:

> When the problem happens, this appears on dmesg:
> [227725.132861] [drm:ironlake_irq_handler [i915]] *ERROR* PCH transcoder 
> B FIFO underrun

Update: I noticed that the problem does not happen when my laptop is on
battery power.


Enrico

-- 
GPG key: 4096R/E7AD5568 2009-05-08 Enrico Zini 


signature.asc
Description: PGP signature

Bug#792746: Upgrading to 4.0.8-1 kernel makes external monitors unusable

[Enrico Zini]

Package: src:linux
Version: 4.5.4-1
Followup-For: Bug #792746

Hello,

thank you for maintaining the Linux kernel in Debian!

I also have this problem. It was better with 4.3, then it became
terrible with 4.4 (visiting graphically intensive sites in Chromium
would turn off the external DisplayPort) and with 4.5, DisplayPort
switches off only when, after a resume, xrandr sets it to the external
monitor's native resolution from the mirror of the laptop's screen that
was used by light-locker..

I'm on a HP EliteBook 2540p, this seems to be the information about the
video card:

00:02.0 VGA compatible controller: Intel Corporation Core Processor Integrated 
Graphics Controller (rev 02) (prog-if 00 [VGA controller])
DeviceName: 32
Subsystem: Hewlett-Packard Company Core Processor Integrated Graphics 
Controller
Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- 
Stepping- SERR- FastB2B- DisINTx+
Status: Cap+ 66MHz- UDF- FastB2B+ ParErr- DEVSEL=fast >TAbort- SERR-  [disabled]
Capabilities: 
Kernel driver in use: i915
Kernel modules: i915

When the problem happens, this appears on dmesg:

[227725.132861] [drm:ironlake_irq_handler [i915]] *ERROR* PCH transcoder B 
FIFO underrun

I currently use this script to reliably turn the external DisplayPort
off and on until it's back on again:

#!/bin/sh
while true
do
PRE=$(dmesg|grep "FIFO underrun" | tail -1)
xrandr --output DP2 --off
sleep 0.3
xrandr --output DP2 --auto $DP2_SETUP
sleep 0.3
POST=$(dmesg|grep "FIFO underrun" | tail -1)
if [ "$PRE" != "$POST" ]
then
dmesg | tail
else
# TODO: once it's tested a few times, send a bug report with this
echo "Should be ok now"
break
fi
done

This seems related: https://bugs.archlinux.org/task/47937


Regards,

Enrico


-- Package-specific info:
** Version:
Linux version 4.5.0-2-amd64 (debian-ker...@lists.debian.org) (gcc version 5.3.1 
20160429 (Debian 5.3.1-17) ) #1 SMP Debian 4.5.3-2 (2016-05-08)

** Command line:
BOOT_IMAGE=/vmlinuz-4.5.0-2-amd64 root=/dev/mapper/viaza-root ro quiet splash

** Tainted: OE (12288)
 * Out-of-tree module has been loaded.
 * Unsigned module has been loaded (currently expected).

** Kernel log:
[227705.634128] microcode: CPU0 microcode updated early to revision 0x4, date = 
2013-06-28
[227705.634171] Enabling non-boot CPUs ...
[227705.653390] x86: Booting SMP configuration:
[227705.653391] smpboot: Booting Node 0 Processor 1 APIC 0x1
[227705.655984]  cache: parent cpu1 should not be sleeping
[227705.656279] CPU1 is up
[227705.673403] smpboot: Booting Node 0 Processor 2 APIC 0x4
[227705.673781] microcode: CPU2 microcode updated early to revision 0x4, date = 
2013-06-28
[227705.676115]  cache: parent cpu2 should not be sleeping
[227705.676414] CPU2 is up
[227705.693441] smpboot: Booting Node 0 Processor 3 APIC 0x5
[227705.696039]  cache: parent cpu3 should not be sleeping
[227705.696349] CPU3 is up
[227705.699068] ACPI: Waking up from system sleep state S3
[227706.214405] acpi LNXPOWER:02: Turning OFF
[227706.214448] acpi LNXPOWER:01: Turning OFF
[227706.214930] yenta_cardbus :44:06.2: enabling device (0006 -> 0007)
[227706.229261] ehci-pci :00:1a.0: System wakeup disabled by ACPI
[227706.229562] ehci-pci :00:1d.0: System wakeup disabled by ACPI
[227706.229641] firewire_ohci :44:06.0: proprietary Ricoh MMC controller 
disabled (via firewire function)
[227706.229643] firewire_ohci :44:06.0: MMC cards are now supported by 
standard SDHCI controller
[227706.233414] PM: noirq resume of devices complete after 18.954 msecs
[227706.235521] PM: early resume of devices complete after 2.086 msecs
[227706.235609] e1000e :00:19.0: System wakeup disabled by ACPI
[227706.235956] sdhci-pci :44:06.1: Will use DMA mode even though HW 
doesn't fully claim to support it.
[227706.236956] sdhci-pci :44:06.1: Will use DMA mode even though HW 
doesn't fully claim to support it.
[227706.242092] sd 0:0:0:0: [sda] Starting disk
[227706.242133] sd 1:0:0:0: [sdc] Starting disk
[227706.369224] tpm_tis 00:01: TPM is disabled/deactivated (0x7)
[227706.369231] rtc_cmos 00:03: System wakeup disabled by ACPI
[227706.465215] usb 1-1.1: reset full-speed USB device number 3 using ehci-pci
[227706.572190] ata3: SATA link down (SStatus 0 SControl 300)
[227706.572473] ata2: SATA link up 3.0 Gbps (SStatus 123 SControl 300)
[227706.572505] ata6: SATA link down (SStatus 0 SControl 300)
[227706.572898] ata2.00: unexpected _GTF length (8)
[227706.572958] ata2.00: supports DRM functions and may not be fully accessible
[227706.575881] ata2.00: disabling queued TRIM support
[227706.579519] ata2.00: unexpected _GTF length (8)
[227706.579592] ata2.00: supports DRM functions and may not be fully accessible
[227706.582499] ata2.00: disabling queued TRIM support
[227706.585716] ata2.00: configured for