You are not logged in.
Pages: 1
The servers are either Dell R630/T630 with Xeon E5-26xx v4 CPUs.
After upgrading to 4.15 kernel(s) the default tty console is not shown. On boot only a few messages (systemd, etc..) is printed but the login prompt is not shown and the server seemd to have freezed no matter what I do with "Alt" + "F?".
However the system did successfully boot and the sshd is on and can be accessed over network. After switching to linux-lts which is 4.14 the issue is resolved --- after the messages the login prompt is out and I can login as usual.
Did this happen to anyone and any solution to the disappeared console? Thanks.
Last edited by wuxb (2018-03-06 21:23:47)
Offline
The issue affects all 4.15 kernels including 4.15.7? Please post a dmesg from a boot under 4.15.7 and one from under 4.14.23.
Edit:
Also the output of lsmod from both 4.14 and 4.15 please.
Last edited by loqs (2018-03-06 22:00:22)
Offline
I'm having exactly this problem with a Thinkmate/SuperMicro server, so it's not just Dell.
Offline
@pgoetz Please post the output of dmesg and lsmod from a boot under 4.15.7 and one from under 4.14.23.
Offline
This appears to be the related error from the dmesg output:
[ 2.961194] clocksource: Switched to clocksource tsc
[ 2.969525] checking generic (dd000000 12c000) vs hw (dd000000 1000000)
[ 2.969528] fb: switching to mgag200drmfb from simple
[ 2.969566] Console: switching to colour dummy device 80x25
[ 2.970045] [drm:mgag200_driver_load [mgag200]] *ERROR* can't reserve VRAM
[ 2.970057] mgag200 0000:09:03.0: Fatal error during GPU init: -6
[ 3.007366] systemd-journald[403]: Received request to flush runtime journal
I'm not running the newest kernel:
[root@www ~]# uname -a
Linux www 4.15.2-2-ARCH #1 SMP PREEMPT Thu Feb 8 18:54:52 UTC 2018 x86_64 GNU/Linux
Unfortunately this machine is a rackmount server, so I hardly ever see the console and am not sure when the issue started.
Offline
@pgoetz Please post the output of dmesg and lsmod from a boot under 4.15.7 and one from under 4.14.23.
This is a production server, so I can't just reboot to 4.14. I can try recreating the problem on another machine, though.
Last edited by pgoetz (2018-03-07 12:14:32)
Offline
Comparing the lsmod output from this server to another SuperMicro-based system with the same video chip, but running the 4.13.12 kernel, it looks like the video driver, mgag200, is being loaded in both cases. Maybe this is only a problem for systems with Matrox MGA video chips?
Server with no console and 4.15.2:
[root@www ~]# lsmod
Module Size Used by
ipt_REJECT 16384 5
nf_reject_ipv4 16384 1 ipt_REJECT
xt_conntrack 16384 6
iptable_filter 16384 1
xt_nat 16384 4
xt_tcpudp 16384 17
iptable_nat 16384 1
nf_conntrack_ipv4 16384 11
nf_defrag_ipv4 16384 1 nf_conntrack_ipv4
nf_nat_ipv4 16384 1 iptable_nat
nf_nat 36864 2 xt_nat,nf_nat_ipv4
nf_conntrack 155648 5 xt_nat,nf_conntrack_ipv4,xt_conntrack,nf_nat_ipv4,nf_nat
libcrc32c 16384 2 nf_conntrack,nf_nat
crc32c_generic 16384 0
iptable_mangle 16384 0
intel_rapl 24576 0
sb_edac 24576 0
x86_pkg_temp_thermal 16384 0
intel_powerclamp 16384 0
coretemp 16384 0
kvm_intel 229376 0
kvm 704512 1 kvm_intel
mgag200 53248 0
ttm 118784 1 mgag200
drm_kms_helper 200704 1 mgag200
irqbypass 16384 1 kvm
crct10dif_pclmul 16384 0
crc32_pclmul 16384 0
ghash_clmulni_intel 16384 0
input_leds 16384 0
pcbc 16384 0
joydev 24576 0
mousedev 24576 0
led_class 16384 1 input_leds
drm 466944 3 mgag200,ttm,drm_kms_helper
aesni_intel 188416 1
ipmi_ssif 32768 0
agpgart 49152 2 ttm,drm
syscopyarea 16384 1 drm_kms_helper
sysfillrect 16384 1 drm_kms_helper
nls_iso8859_1 16384 1
aes_x86_64 20480 1 aesni_intel
nls_cp437 20480 1
crypto_simd 16384 1 aesni_intel
vfat 20480 1
fat 77824 1 vfat
glue_helper 16384 1 aesni_intel
igb 245760 0
ptp 20480 1 igb
pps_core 20480 1 ptp
cryptd 28672 3 crypto_simd,ghash_clmulni_intel,aesni_intel
sysimgblt 16384 1 drm_kms_helper
mxm_wmi 16384 0
ioatdma 61440 0
fb_sys_fops 16384 1 drm_kms_helper
i2c_algo_bit 16384 2 mgag200,igb
hid_generic 16384 0
intel_cstate 16384 0
sr_mod 28672 0
iTCO_wdt 16384 0
iTCO_vendor_support 16384 1 iTCO_wdt
mei_me 45056 0
intel_uncore 131072 0
shpchp 40960 0
intel_rapl_perf 16384 0
dca 16384 2 ioatdma,igb
mei 106496 1 mei_me
i2c_i801 32768 0
pcspkr 16384 0
lpc_ich 28672 0
cdrom 69632 1 sr_mod
ipmi_si 65536 0
ipmi_devintf 20480 0
ipmi_msghandler 53248 3 ipmi_ssif,ipmi_devintf,ipmi_si
rtc_cmos 24576 1
wmi 28672 1 mxm_wmi
evdev 20480 5
usbhid 57344 0
mac_hid 16384 0
hid 131072 2 hid_generic,usbhid
sch_fq_codel 20480 17
ip_tables 28672 3 iptable_mangle,iptable_filter,iptable_nat
x_tables 45056 7 ipt_REJECT,xt_nat,iptable_mangle,ip_tables,iptable_filter,xt_tcpudp,xt_conntrack
ext4 712704 2
crc16 16384 1 ext4
mbcache 16384 1 ext4
jbd2 122880 1 ext4
fscrypto 32768 1 ext4
sd_mod 61440 6
ahci 40960 0
libahci 40960 1 ahci
crc32c_intel 24576 1
libata 278528 2 ahci,libahci
megaraid_sas 163840 4
ehci_pci 16384 0
ehci_hcd 94208 1 ehci_pci
scsi_mod 258048 4 sd_mod,megaraid_sas,libata,sr_mod
usbcore 286720 3 usbhid,ehci_hcd,ehci_pci
usb_common 16384 1 usbcore
Server with console running 4.13.12
[root@adam ~]# lsmod
Module Size Used by
nls_iso8859_1 16384 1
nls_cp437 20480 1
vfat 20480 1
fat 65536 1 vfat
iTCO_wdt 16384 0
ipmi_ssif 24576 0
iTCO_vendor_support 16384 1 iTCO_wdt
intel_rapl 20480 0
x86_pkg_temp_thermal 16384 0
intel_powerclamp 16384 0
coretemp 16384 0
kvm_intel 192512 0
kvm 516096 1 kvm_intel
irqbypass 16384 1 kvm
crct10dif_pclmul 16384 0
crc32_pclmul 16384 0
ghash_clmulni_intel 16384 0
pcbc 16384 0
aesni_intel 184320 0
aes_x86_64 20480 1 aesni_intel
crypto_simd 16384 1 aesni_intel
glue_helper 16384 1 aesni_intel
cryptd 20480 3 crypto_simd,ghash_clmulni_intel,aesni_intel
intel_cstate 16384 0
intel_rapl_perf 16384 0
pcspkr 16384 0
mgag200 40960 1
ttm 81920 1 mgag200
drm_kms_helper 131072 1 mgag200
drm 303104 4 mgag200,ttm,drm_kms_helper
agpgart 36864 2 ttm,drm
syscopyarea 16384 1 drm_kms_helper
sysfillrect 16384 1 drm_kms_helper
igb 200704 0
sysimgblt 16384 1 drm_kms_helper
mei_me 36864 0
ptp 20480 1 igb
lpc_ich 24576 0
mei 81920 1 mei_me
fb_sys_fops 16384 1 drm_kms_helper
pps_core 20480 1 ptp
i2c_algo_bit 16384 2 mgag200,igb
i2c_i801 24576 0
ioatdma 49152 0
dca 16384 2 ioatdma,igb
shpchp 32768 0
ipmi_si 49152 0
ipmi_devintf 20480 0
ipmi_msghandler 40960 3 ipmi_ssif,ipmi_devintf,ipmi_si
tpm_tis 16384 0
tpm_tis_core 20480 1 tpm_tis
wmi 20480 0
tpm 49152 2 tpm_tis,tpm_tis_core
joydev 20480 0
mousedev 20480 0
button 16384 0
evdev 24576 5
input_leds 16384 0
led_class 16384 1 input_leds
mac_hid 16384 0
sch_fq_codel 20480 9
ip_tables 24576 0
x_tables 32768 1 ip_tables
ext4 557056 1
crc16 16384 1 ext4
mbcache 16384 1 ext4
jbd2 90112 1 ext4
fscrypto 24576 1 ext4
hid_generic 16384 0
usbhid 45056 0
hid 114688 2 hid_generic,usbhid
raid1 40960 2
md_mod 126976 3 raid1
ses 20480 0
enclosure 16384 1 ses
sd_mod 49152 8
scsi_transport_sas 36864 1 ses
crc32c_intel 24576 0
megaraid_sas 135168 0
ahci 36864 6
libahci 28672 1 ahci
ehci_pci 16384 0
libata 208896 2 ahci,libahci
ehci_hcd 73728 1 ehci_pci
usbcore 208896 3 usbhid,ehci_hcd,ehci_pci
scsi_mod 155648 5 sd_mod,megaraid_sas,ses,scsi_transport_sas,libata
usb_common 16384 1 usbcore
Offline
On a machine you can recreate the issue on can you try linux 4.15.1-2 from the ALA.
Edit: Or any linux > 4.15.3-1 although linux > 4.15.5-1 would be preferred.
Last edited by loqs (2018-03-07 18:20:07)
Offline
I have another server in the same rack with the same graphics chip that hasn't been updated yet. The simplest thing to do is just run updates on that machine and see if it suddenly loses console, too.
Offline
I resolved my issue by upgrading from the 4.15.2 to the 4.15.7 kernel. First time I've ever had a kernel update result in a failure to load console.
Offline
You just picked an unfortunate kernel to update to if you do not do regular updates. Only 4.15.1-3 to 4.15-3-1 inclusive had CONFIG_X86_SYSFB=y
Edit:
spelling inclusive not includsive
Last edited by loqs (2018-03-09 16:28:48)
Offline
Only 4.15.1-3 to 4.15-3-1 inclusive had CONFIG_X86_SYSFB=y
Thanks for that helpful explanatory note. Do you know why this option was changed for just that kernel sequence?
Offline
The commit log does not have a reason for the initial inclusion of the option https://git.archlinux.org/svntogit/pack … ages/linux the removal could be related to "Pick some config options from Ubuntu"
which incidentally caused the same issue but for systems with nvidia GPU's but not using the nvidia driver which in turn had to be reverted in "Disable non-generic, non-virtual FB drivers again".
Offline
Pages: 1