6:15 a.m.
Dear fellow Fedorans,
Fedora 34 on my i7 with 32G RAM and nvidia RTX2060 card takes minutes to
boot, and when it finally does there are a number of "something went
wrong" notifications. How best can I diagnose and fix this so that it
boots quickly and without errors?
CPU: Quad Core Intel Core i7-3770K (-MT MCP-) speed/min/max:
4324/1600/4400 MHz Kernel: 5.12.11-300.fc34.x86_64 x86_64
Up: 2h 30m Mem: 8717.9/31785.3 MiB (27.4%) Storage: 25.69 TiB (59.7%
used) Procs: 391 Shell: Bash inxi: 3.3.03
The top part of systemd-analyze-blame is
1min 23.232s plymouth-quit-wait.service
53.077s cs-firewall-bouncer.service
52.219s dovecot.service
26.525s crowdsec.service
26.075s libvirtd.service
25.910s postfix.service
25.888s vmware.service
25.870s nfs-server.service
8.275s network.service
4.279s abrtd.service
4.111s smartd.service
1.921s systemd-udev-settle.service
1.355s lvm2-monitor.service
1.309s user(a)1006.service
1.188s
systemd-fsck@dev-disk-by\x2duuid-d91d02e3\x2dce79\x2d4aa8\x2d9446\x2df531c05ca7a7.service
1.145s udisks2.service
711ms akmods.service
602ms initrd-switch-root.service
Many thanks,
Chris R.
7:11 a.m.
On Thu, Jun 24, 2021 at 12:15:48PM +0100, Christopher Ross wrote:
Fedora 34 on my i7 with 32G RAM and nvidia RTX2060 card takes minutes
to
boot, and when it finally does there are a number of "something went wrong"
notifications. How best can I diagnose and fix this so that it boots quickly
and without errors?
Do you have the RPMFusion nvidia packages installed? Are you using
any 3rd-party nvidia drivers? Or are you using the nouveau driver?
The nvidia driver might be compiling on boot (dkms) which takes a long
time, and if it fails, will cause GL issues that can break 'nautilus',
and if 'nautilus' crashes, the GNOME session will do the 'Something
went wrong' alert.
The top part of systemd-analyze-blame is
1min 23.232s plymouth-quit-wait.service
53.077s cs-firewall-bouncer.service
52.219s dovecot.service
26.525s crowdsec.service
It appears you're using some sort of 3rd-party firewall driver called
'crowdsec'. Is that the problem? It does seem to be up there,
although it could be waiting for something else to start.
--
Jonathan Billings <billings(a)negate.org>
7:34 a.m.
On 24/06/2021 13:11, Jonathan Billings wrote:
On Thu, Jun 24, 2021 at 12:15:48PM +0100, Christopher Ross wrote:
> Fedora 34 on my i7 with 32G RAM and nvidia RTX2060 card takes minutes to
> boot, and when it finally does there are a number of "something went
wrong"
> notifications. How best can I diagnose and fix this so that it boots quickly
> and without errors?
Do you have the RPMFusion nvidia packages installed? Are you using
any 3rd-party nvidia drivers? Or are you using the nouveau driver?
Yes, RPMFusion is enabled and nvidia drivers installed.
The nvidia driver might be compiling on boot (dkms) which takes a
long
time, and if it fails, will cause GL issues that can break 'nautilus',
and if 'nautilus' crashes, the GNOME session will do the 'Something
went wrong' alert.
It happens every boot, not just when there is a kernel or driver update.
In that instance it should not need to recompile.
I should have mentioned that I'm running KDE not Gnome, but at this
stage it hasn't run. The 3½ minutes is the time it takes from power on
to bring up the GDM login screen.
I can't remember the exact message from abrtd. I good follow up question
is probably how can I get more information about the oopses?
Looking in /var/spool/abrt/oops-2021-06-24-09:48:20-3565-0/dmesg
It does seem the first oops was nvidia
...
[ 14.920735] intel_rapl_common: RAPL package-0 domain package
locked by BIOS
[ 14.956559] pktcdvd: pktcdvd0: writer mapped to sr0
[ 15.073374] zram0: detected capacity change from 0 to 16777216
[ 15.095100] Adding 8388604k swap on /dev/zram0. Priority:100
extents:1 across:8388604k SSFS
[ 15.374420] [drm] Initialized nvidia-drm 0.0.0 20160202 for
0000:01:00.0 on minor 1
[ 15.684579] nvidia-gpu 0000:01:00.3: i2c timeout error e0000000
[ 15.684583] ucsi_ccg 8-0008: i2c_transfer failed -110
[ 15.684585] ucsi_ccg 8-0008: ucsi_ccg_init failed - -110
[ 15.684590] ucsi_ccg: probe of 8-0008 failed with error -110
[ 40.291071] watchdog: BUG: soft lockup - CPU#3 stuck for 26s!
[plymouthd:445]
[ 40.291074] Modules linked in: intel_rapl_msr intel_rapl_common
at24 mei_hdcp iTCO_wdt intel_pmc_bxt iTCO_vendor_support ucsi_ccg
typec_ucsi typec pktcdvd x86_pkg_temp_thermal intel_powerclamp
coretemp kvm_intel kvm irqbypass rapl intel_cstate intel_uncore
raid0 eeepc_wmi asus_wmi sparse_keymap rfkill wmi_bmof pcspkr
snd_hda_codec_realtek snd_hda_codec_generic ledtrig_audio
nvidia_drm(POE) joydev snd_hda_codec_hdmi nvidia_modeset(POE)
snd_hda_intel i2c_i801 snd_intel_dspcfg i2c_smbus snd_intel_sdw_acpi
snd_usb_audio apple_mfi_fastcharge snd_hda_codec nvidia_uvm(POE)
snd_usbmidi_lib snd_hda_core snd_rawmidi mc snd_hwdep snd_seq
snd_seq_device lpc_ich snd_pcm nvidia(POE) mei_me snd_timer snd mei
i2c_nvidia_gpu soundcore nfsd auth_rpcgss nfs_acl lockd grace
binfmt_misc sunrpc nfs_ssc zram ip_tables i915 crct10dif_pclmul
crc32_pclmul crc32c_intel cdc_mbim cdc_wdm e1000e mxm_wmi
i2c_algo_bit ghash_clmulni_intel drm_kms_helper cec cdc_ncm
cdc_ether drm usbnet mii wmi video fuse
[ 40.291104] CPU: 3 PID: 445 Comm: plymouthd Tainted: P
OE 5.12.11-300.fc34.x86_64 #1
[ 40.291106] Hardware name: System manufacturer System Product
Name/MAXIMUS V FORMULA, BIOS 1903 08/19/2013
[ 40.291107] RIP: 0010:os_io_read_dword+0x8/0x10 [nvidia]
[ 40.291311] Code: 00 00 0f 1f 44 00 00 89 fa ec c3 0f 1f 80 00 00
00 00 0f 1f 44 00 00 89 fa 66 ed c3 66 0f 1f 44 00 00 0f 1f 44 00 00
89 fa ed <c3> 0f 1f 80 00 00 00 00 0f 1f 44 00 00 48 85 ff 75 0c 48
8b 05 c7
[ 40.291312] RSP: 0018:ffffaeb9c05cf7d8 EFLAGS: 00000202
[ 40.291313] RAX: 0000000003763580 RBX: 00000000000017d3 RCX:
00000000000c0000
[ 40.291314] RDX: 000000000000e00c RSI: 00000000000c40e4 RDI:
000000000000e00c
[ 40.291314] RBP: ffff8c7220e12b10 R08: ffffffffc2aa5380 R09:
0000000000000282
[ 40.291315] R10: 0000000000000202 R11: 0000000000000040 R12:
ffff8c7220e12b3c
[ 40.291316] R13: ffff8c7220e12b38 R14: 000000000000c000 R15:
000000000000c000
[ 40.291316] FS: 00007f27013a4800(0000) GS:ffff8c790fec0000(0000)
knlGS:0000000000000000
[ 40.291317] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[ 40.291318] CR2: 00007ffca2940078 CR3: 000000010a834001 CR4:
00000000001706e0
[ 40.291319] Call Trace:
[ 40.291321] _nv041000rm+0x4c/0x70 [nvidia]
[ 40.291545] ? _nv040998rm+0x30/0x30 [nvidia]
[ 40.291768] ? _nv000834rm+0x4f/0x130 [nvidia]
...
> The top part of systemd-analyze-blame is
>
> 1min 23.232s plymouth-quit-wait.service
> 53.077s cs-firewall-bouncer.service
> 52.219s dovecot.service
> 26.525s crowdsec.service
It appears you're using some sort of 3rd-party firewall driver called
'crowdsec'. Is that the problem? It does seem to be up there,
although it could be waiting for something else to start.
Crowdsec is the modern replacement for fail2ban
https://crowdsec.net/
But disabling it doesn't help. It's presumably waiting on the network.
Many thanks,
Chris R.
8:05 a.m.
On Thu, Jun 24, 2021 at 01:34:27PM +0100, Christopher Ross wrote:
I should have mentioned that I'm running KDE not Gnome, but at
this stage it
hasn't run. The 3½ minutes is the time it takes from power on to bring up
the GDM login screen.
I can't remember the exact message from abrtd. I good follow up question is
probably how can I get more information about the oopses?
Looking in /var/spool/abrt/oops-2021-06-24-09:48:20-3565-0/dmesg
It does seem the first oops was nvidia
...
[ 14.920735] intel_rapl_common: RAPL package-0 domain package
locked by BIOS
[ 14.956559] pktcdvd: pktcdvd0: writer mapped to sr0
[ 15.073374] zram0: detected capacity change from 0 to 16777216
[ 15.095100] Adding 8388604k swap on /dev/zram0. Priority:100
extents:1 across:8388604k SSFS
[ 15.374420] [drm] Initialized nvidia-drm 0.0.0 20160202 for
0000:01:00.0 on minor 1
[ 15.684579] nvidia-gpu 0000:01:00.3: i2c timeout error e0000000
[ 15.684583] ucsi_ccg 8-0008: i2c_transfer failed -110
[ 15.684585] ucsi_ccg 8-0008: ucsi_ccg_init failed - -110
[ 15.684590] ucsi_ccg: probe of 8-0008 failed with error -110
[ 40.291071] watchdog: BUG: soft lockup - CPU#3 stuck for 26s!
[plymouthd:445]
[ 40.291074] Modules linked in: intel_rapl_msr intel_rapl_common
at24 mei_hdcp iTCO_wdt intel_pmc_bxt iTCO_vendor_support ucsi_ccg
typec_ucsi typec pktcdvd x86_pkg_temp_thermal intel_powerclamp
coretemp kvm_intel kvm irqbypass rapl intel_cstate intel_uncore
raid0 eeepc_wmi asus_wmi sparse_keymap rfkill wmi_bmof pcspkr
snd_hda_codec_realtek snd_hda_codec_generic ledtrig_audio
nvidia_drm(POE) joydev snd_hda_codec_hdmi nvidia_modeset(POE)
snd_hda_intel i2c_i801 snd_intel_dspcfg i2c_smbus snd_intel_sdw_acpi
snd_usb_audio apple_mfi_fastcharge snd_hda_codec nvidia_uvm(POE)
snd_usbmidi_lib snd_hda_core snd_rawmidi mc snd_hwdep snd_seq
snd_seq_device lpc_ich snd_pcm nvidia(POE) mei_me snd_timer snd mei
i2c_nvidia_gpu soundcore nfsd auth_rpcgss nfs_acl lockd grace
binfmt_misc sunrpc nfs_ssc zram ip_tables i915 crct10dif_pclmul
crc32_pclmul crc32c_intel cdc_mbim cdc_wdm e1000e mxm_wmi
i2c_algo_bit ghash_clmulni_intel drm_kms_helper cec cdc_ncm
cdc_ether drm usbnet mii wmi video fuse
[ 40.291104] CPU: 3 PID: 445 Comm: plymouthd Tainted: P
OE 5.12.11-300.fc34.x86_64 #1
[ 40.291106] Hardware name: System manufacturer System Product
Name/MAXIMUS V FORMULA, BIOS 1903 08/19/2013
[ 40.291107] RIP: 0010:os_io_read_dword+0x8/0x10 [nvidia]
[ 40.291311] Code: 00 00 0f 1f 44 00 00 89 fa ec c3 0f 1f 80 00 00
00 00 0f 1f 44 00 00 89 fa 66 ed c3 66 0f 1f 44 00 00 0f 1f 44 00 00
89 fa ed <c3> 0f 1f 80 00 00 00 00 0f 1f 44 00 00 48 85 ff 75 0c 48
8b 05 c7
[ 40.291312] RSP: 0018:ffffaeb9c05cf7d8 EFLAGS: 00000202
[ 40.291313] RAX: 0000000003763580 RBX: 00000000000017d3 RCX:
00000000000c0000
[ 40.291314] RDX: 000000000000e00c RSI: 00000000000c40e4 RDI:
000000000000e00c
[ 40.291314] RBP: ffff8c7220e12b10 R08: ffffffffc2aa5380 R09:
0000000000000282
[ 40.291315] R10: 0000000000000202 R11: 0000000000000040 R12:
ffff8c7220e12b3c
[ 40.291316] R13: ffff8c7220e12b38 R14: 000000000000c000 R15:
000000000000c000
[ 40.291316] FS: 00007f27013a4800(0000) GS:ffff8c790fec0000(0000)
knlGS:0000000000000000
[ 40.291317] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[ 40.291318] CR2: 00007ffca2940078 CR3: 000000010a834001 CR4:
00000000001706e0
[ 40.291319] Call Trace:
[ 40.291321] _nv041000rm+0x4c/0x70 [nvidia]
[ 40.291545] ? _nv040998rm+0x30/0x30 [nvidia]
[ 40.291768] ? _nv000834rm+0x4f/0x130 [nvidia]
I agree, it looks like a nvidia-related issue. What kernel arguments
do you have (look at /proc/cmdline or edit the GRUB entry when you're
booting). Also, it looks like the nvidia modeset driver was running,
which might be having problems with plymouth?
You could try booting without the 'rhgb quiet' arguments in the kernel
line in GRUB, so you can see if disabling plymouth helps, and it will
show you what is actually happening during boot.
I was under the impression that GDM + KDE didn't work well, maybe try
using a KDE-friendly login manager like sddm?
--
Jonathan Billings <billings(a)negate.org>
8:21 a.m.
On 24/06/2021 14:05, Jonathan Billings wrote:
On Thu, Jun 24, 2021 at 01:34:27PM +0100, Christopher Ross wrote:
> I should have mentioned that I'm running KDE not Gnome, but at this stage it
> hasn't run. The 3½ minutes is the time it takes from power on to bring up
> the GDM login screen.
>
> I can't remember the exact message from abrtd. I good follow up question is
> probably how can I get more information about the oopses?
>
> Looking in /var/spool/abrt/oops-2021-06-24-09:48:20-3565-0/dmesg
>
> It does seem the first oops was nvidia
>
> ...
> [ 14.920735] intel_rapl_common: RAPL package-0 domain package
> locked by BIOS
> [ 14.956559] pktcdvd: pktcdvd0: writer mapped to sr0
> [ 15.073374] zram0: detected capacity change from 0 to 16777216
> [ 15.095100] Adding 8388604k swap on /dev/zram0. Priority:100
> extents:1 across:8388604k SSFS
> [ 15.374420] [drm] Initialized nvidia-drm 0.0.0 20160202 for
> 0000:01:00.0 on minor 1
> [ 15.684579] nvidia-gpu 0000:01:00.3: i2c timeout error e0000000
> [ 15.684583] ucsi_ccg 8-0008: i2c_transfer failed -110
> [ 15.684585] ucsi_ccg 8-0008: ucsi_ccg_init failed - -110
> [ 15.684590] ucsi_ccg: probe of 8-0008 failed with error -110
> [ 40.291071] watchdog: BUG: soft lockup - CPU#3 stuck for 26s!
> [plymouthd:445]
> [ 40.291074] Modules linked in: intel_rapl_msr intel_rapl_common
> at24 mei_hdcp iTCO_wdt intel_pmc_bxt iTCO_vendor_support ucsi_ccg
> typec_ucsi typec pktcdvd x86_pkg_temp_thermal intel_powerclamp
> coretemp kvm_intel kvm irqbypass rapl intel_cstate intel_uncore
> raid0 eeepc_wmi asus_wmi sparse_keymap rfkill wmi_bmof pcspkr
> snd_hda_codec_realtek snd_hda_codec_generic ledtrig_audio
> nvidia_drm(POE) joydev snd_hda_codec_hdmi nvidia_modeset(POE)
> snd_hda_intel i2c_i801 snd_intel_dspcfg i2c_smbus snd_intel_sdw_acpi
> snd_usb_audio apple_mfi_fastcharge snd_hda_codec nvidia_uvm(POE)
> snd_usbmidi_lib snd_hda_core snd_rawmidi mc snd_hwdep snd_seq
> snd_seq_device lpc_ich snd_pcm nvidia(POE) mei_me snd_timer snd mei
> i2c_nvidia_gpu soundcore nfsd auth_rpcgss nfs_acl lockd grace
> binfmt_misc sunrpc nfs_ssc zram ip_tables i915 crct10dif_pclmul
> crc32_pclmul crc32c_intel cdc_mbim cdc_wdm e1000e mxm_wmi
> i2c_algo_bit ghash_clmulni_intel drm_kms_helper cec cdc_ncm
> cdc_ether drm usbnet mii wmi video fuse
> [ 40.291104] CPU: 3 PID: 445 Comm: plymouthd Tainted: P
> OE 5.12.11-300.fc34.x86_64 #1
> [ 40.291106] Hardware name: System manufacturer System Product
> Name/MAXIMUS V FORMULA, BIOS 1903 08/19/2013
> [ 40.291107] RIP: 0010:os_io_read_dword+0x8/0x10 [nvidia]
> [ 40.291311] Code: 00 00 0f 1f 44 00 00 89 fa ec c3 0f 1f 80 00 00
> 00 00 0f 1f 44 00 00 89 fa 66 ed c3 66 0f 1f 44 00 00 0f 1f 44 00 00
> 89 fa ed <c3> 0f 1f 80 00 00 00 00 0f 1f 44 00 00 48 85 ff 75 0c 48
> 8b 05 c7
> [ 40.291312] RSP: 0018:ffffaeb9c05cf7d8 EFLAGS: 00000202
> [ 40.291313] RAX: 0000000003763580 RBX: 00000000000017d3 RCX:
> 00000000000c0000
> [ 40.291314] RDX: 000000000000e00c RSI: 00000000000c40e4 RDI:
> 000000000000e00c
> [ 40.291314] RBP: ffff8c7220e12b10 R08: ffffffffc2aa5380 R09:
> 0000000000000282
> [ 40.291315] R10: 0000000000000202 R11: 0000000000000040 R12:
> ffff8c7220e12b3c
> [ 40.291316] R13: ffff8c7220e12b38 R14: 000000000000c000 R15:
> 000000000000c000
> [ 40.291316] FS: 00007f27013a4800(0000) GS:ffff8c790fec0000(0000)
> knlGS:0000000000000000
> [ 40.291317] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
> [ 40.291318] CR2: 00007ffca2940078 CR3: 000000010a834001 CR4:
> 00000000001706e0
> [ 40.291319] Call Trace:
> [ 40.291321] _nv041000rm+0x4c/0x70 [nvidia]
> [ 40.291545] ? _nv040998rm+0x30/0x30 [nvidia]
> [ 40.291768] ? _nv000834rm+0x4f/0x130 [nvidia]
I agree, it looks like a nvidia-related issue. What kernel arguments
do you have (look at /proc/cmdline or edit the GRUB entry when you're
booting). Also, it looks like the nvidia modeset driver was running,
which might be having problems with plymouth?
You could try booting without the 'rhgb quiet' arguments in the kernel
line in GRUB, so you can see if disabling plymouth helps, and it will
show you what is actually happening during boot.
I was under the impression that GDM + KDE didn't work well, maybe try
using a KDE-friendly login manager like sddm?
I only recently changed from sddm to gdm because the "switch user"
functionality has been removed from sddm in Fedora 34. We do share this
computer.
That change didn't fix the boot issue either.
Regards,
Chris R.
10:03 a.m.
On Thu, 2021-06-24 at 14:21 +0100, Christopher Ross wrote:
> I was under the impression that GDM + KDE didn't work well,
maybe
> try
> using a KDE-friendly login manager like sddm?
I only recently changed from sddm to gdm because the "switch user"
functionality has been removed from sddm in Fedora 34.
Same here.
poc
11:24 a.m.
On 6/24/21 5:15 AM, Christopher Ross wrote:
The top part of systemd-analyze-blame is
1min 23.232s plymouth-quit-wait.service
53.077s cs-firewall-bouncer.service
52.219s dovecot.service
26.525s crowdsec.service
26.075s libvirtd.service
25.910s postfix.service
25.888s vmware.service
25.870s nfs-server.service
This is just a guess, but it looks to me like the firewall-bouncer
service is causing the delay and the others are just waiting for it to
become available.
1030
days inactive
1030
days old
6 comments
4 participants
participants (4)
-
Christopher Ross -
Joe Zeff -
Jonathan Billings -
Patrick O'Callaghan