Acer Swift Go 14 compatibility issues

Hello !
I have some issues with my Acer Swift Go 14 (model 72)

First time using Fedora, seems really great ! :slight_smile:

First one is the having no sound, but was solved using [BUG] Acer Swift Go 14 - No sound card detected - "No SoundWire machine driver found" · Issue #4923 · thesofproject/linux · GitHub (how to ask for a new alsa-conf version is a side question for this topic)

But the second one is related to boot times, nearly 2 minutes. I struggle reading and understanding all the boot time analysis.
Note that I added a second NVME SSD to my laptop on which Fedora is installed.
Note that sometimes, not always, udisk service does not start…

Here is what I tried

vincent@fedora:~$ systemd-analyze blame
1min 1.819s sys-module-fuse.device
1min 1.779s dev-tpm0.device
1min 1.779s sys-devices-LNXSYSTM:00-LNXSYBUS:00-INTC7001:00-tpm-tpm0.device
1min 1.771s dev-ttyS3.device
1min 1.771s sys-devices-platform-serial8250-serial8250:0-serial8250:0.3-tty-tty>
1min 1.770s dev-ttyS2.device
1min 1.770s sys-devices-platform-serial8250-serial8250:0-serial8250:0.2-tty-tty>
1min 1.767s sys-devices-platform-serial8250-serial8250:0-serial8250:0.0-tty-tty>
1min 1.767s dev-ttyS0.device
1min 1.766s sys-devices-platform-serial8250-serial8250:0-serial8250:0.1-tty-tty>
1min 1.766s dev-ttyS1.device
1min 1.758s dev-tpmrm0.device
1min 1.758s sys-devices-LNXSYSTM:00-LNXSYBUS:00-INTC7001:00-tpmrm-tpmrm0.device
1min 1.750s sys-module-configfs.device
    38.237s fwupd.service
     6.368s NetworkManager-wait-online.service
     1.916s plymouth-quit-wait.service
     1.134s dev-disk-by\x2did-nvme\x2dHFS512GEJ9X110N_5YC9N032511709O1J_1\x2dpa>
     1.134s dev-disk-by\x2dpath-pci\x2d0000:00:0e.0\x2dpci\x2d10000:e1:00.0\x2d>
     1.134s dev-nvme1n1p3.device

vincent@fedora:~$ systemd-analyze critical-chain  
The time when unit became active or started is printed after the "@" character.
The time the unit took to start is printed after the "+" character.

graphical.target @1min 5.262s
└─multi-user.target @11.784s
  └─plymouth-quit-wait.service @9.866s +1.916s
    └─systemd-user-sessions.service @9.848s +16ms
      └─remote-fs.target @9.846s
        └─remote-fs-pre.target @3.524s
          └─nfs-client.target @3.524s
            └─gssproxy.service @3.504s +20ms
              └─network.target @3.501s
                └─wpa_supplicant.service @3.478s +22ms
                  └─basic.target @2.535s
                    └─dbus-broker.service @2.503s +30ms
                      └─dbus.socket @2.496s
                        └─sysinit.target @2.493s
                          └─systemd-resolved.service @1.563s +929ms
                            └─run-credentials-systemd\x2dresolved.service.mount>

Some things I saw in dmesg

[    2.408583] fbcon: Deferring console take-over
[    2.408584] i915 0000:00:02.0: [drm] fb0: i915drmfb frame buffer device
[    2.525972] i915 0000:00:02.0: [drm] GT1: Loaded GSC firmware i915/mtl_gsc_1.bin (cv1.0, r102.0.10.1878, svn 1)
[    2.546019] i915 0000:00:02.0: [drm] GT1: HuC: authenticated for all workloads
[    9.840145] fbcon: Taking over console
[    9.857618] Console: switching to colour frame buffer device 360x112
[   22.546610] i915 0000:00:02.0: [drm] *ERROR* GT1: GSC proxy component didn't bind within the expected timeout
[   22.546624] i915 0000:00:02.0: [drm] *ERROR* GT1: GSC proxy handler failed to init
[   62.533401] nvme nvme0: I/O tag 20 (2014) QID 0 timeout, completion polled
[   62.538036] nvme nvme0: 8/0/0 default/read/poll queues
[   62.541366]  nvme0n1: p1 p2 p3
[   62.631692] BTRFS: device label fedora devid 1 transid 141 /dev/nvme0n1p3 (259:8) scanned by mount (687)
[   62.684385] BTRFS info (device nvme0n1p3): first mount of filesystem 0a14acb6-d4fc-443d-be2f-bc943781a1a4
[   62.684410] BTRFS info (device nvme0n1p3): using crc32c (crc32c-intel) checksum algorithm
[   62.684417] BTRFS info (device nvme0n1p3): using free-space-tree
[   62.819676] systemd-journald[403]: Received SIGTERM from PID 1 (systemd).
[   62.859478] audit: type=1404 audit(1728826326.311:5): enforcing=1 old_enforcing=0 auid=4294967295 ses=4294967295 enabled=1 old-enabled=1 lsm=selinux res=1

First is the ‘ERROR GT1: GSC proxy handler failed to init’ which seems to take 20 seconds and then nvme timeout on which Fedora is installed.

Can someone help troubleshooting ? Thanks :slight_smile:

1 Like

Here:

% systemd-analyze critical-chain             
The time when unit became active or started is printed after the "@" character.
The time the unit took to start is printed after the "+" character.

graphical.target @36.764s
└─power-profiles-daemon.service @36.677s +86ms
  └─multi-user.target @36.673s
[...]

Your system reached multi-user.target in under 12 seconds, but took nearly a minute more to reach graphical.target where mine took under 1/10th of a second. The systemd-analyze blame starts with a long section of devices, all with about the same time. This suggests they were all waiting on the same component, probably mass storage or network.
You have:

1min 1.750s sys-module-configfs.device
    38.237s fwupd.service
     6.368s NetworkManager-wait-online.service

Sometimes systemctl status <service> will have some added details. For example, I see:

% systemctl status fwupd.service
● fwupd.service - Firmware update daemon
     Loaded: loaded (/usr/lib/systemd/system/fwupd.service; static)
    Drop-In: /usr/lib/systemd/system/service.d
             └─10-timeout-abort.conf
     Active: active (running) since Sat 2024-10-12 22:01:24 ADT; 11h ago
 Invocation: 34efdd6848dd4034a5d0cb2a0b85910a
       Docs: https://fwupd.org/
   Main PID: 3493 (fwupd)
      Tasks: 6 (limit: 18933)
     Memory: 23.9M (peak: 24.6M)
        CPU: 1.182s
     CGroup: /system.slice/fwupd.service
             └─3493 /usr/libexec/fwupd/fwupd

Oct 12 22:01:23 dormarth systemd[1]: Starting fwupd.service - Firmware update daemon...
Oct 12 22:01:24 dormarth fwupd[3493]: 01:01:24.439 FuPluginIntelMe      failed to get public key using /fpf/OemCred: generic failure [0xb]
Oct 12 22:01:24 dormarth fwupd[3493]: 01:01:24.642 FuMain               Daemon ready for requests (locale en_CA.UTF-8)
Oct 12 22:01:24 dormarth systemd[1]: Started fwupd.service - Firmware update daemon.

This line (along with many others) is truncated. One way to avoid this is to add |cat at the end of the command line when preparing a post.

Thanks a lot for your answers !

TPM is not the suspect here, deactivating it, the services do not try to load but it does not change anything.

Thought for a while that was fwupd.service, as an error was shown on systemctl status fwupd.service, but this is not the case.

My next guess is that the udisk2 service / my SSDs are the main suspect. Because of the gap in dmesg (see first message), and the fact that udisk2 timeouts at boot every time:

oct. 13 17:27:02 fedora systemd[1]: Starting udisks2.service - Disk Manager...
oct. 13 17:27:02 fedora udisksd[1163]: udisks daemon version 2.10.1 starting
oct. 13 17:27:47 fedora systemd[1]: udisks2.service: start operation timed out. Terminating.
oct. 13 17:28:05 fedora udisksd[1163]: udisks daemon version 2.10.1 exiting
oct. 13 17:28:05 fedora systemd[1]: udisks2.service: Failed with result 'timeout'.
oct. 13 17:28:05 fedora systemd[1]: Failed to start udisks2.service - Disk Manager.
oct. 13 17:28:05 fedora systemd[1]: Starting udisks2.service - Disk Manager...
oct. 13 17:28:05 fedora udisksd[3453]: udisks daemon version 2.10.1 starting
oct. 13 17:28:05 fedora systemd[1]: Started udisks2.service - Disk Manager.
oct. 13 17:28:05 fedora udisksd[3453]: Acquired the name org.freedesktop.UDisks2 on the system message bus

Next is to find why :upside_down_face:

For the record:

vincent@fedora:~$ sudo fdisk -l
[sudo] Mot de passe de vincent : 
Disk /dev/nvme1n1: 476,94 GiB, 512110190592 bytes, 1000215216 sectors
Disk model: HFS512GEJ9X110N                         
Units: sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disklabel type: gpt
Disk identifier: AB6979A2-674F-4AFB-888D-F7C2865E57FC

Device             Start        End   Sectors   Size Type
/dev/nvme1n1p1      2048     534527    532480   260M EFI System
/dev/nvme1n1p2    534528     567295     32768    16M Microsoft reserved
/dev/nvme1n1p3    567296  998117375 997550080 475,7G Microsoft basic data
/dev/nvme1n1p4 998117376 1000214527   2097152     1G Windows recovery environment


Disk /dev/nvme0n1: 953,87 GiB, 1024209543168 bytes, 2000409264 sectors
Disk model: KXG50ZNV1T02 NVMe TOSHIBA 1024GB        
Units: sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disklabel type: gpt
Disk identifier: 6CADA518-B655-44DE-9A89-E46CEFD6B2F5

Device           Start        End    Sectors   Size Type
/dev/nvme0n1p1    2048    1230847    1228800   600M EFI System
/dev/nvme0n1p2 1230848    3327999    2097152     1G Linux extended boot
/dev/nvme0n1p3 3328000 2000408575 1997080576 952,3G Linux filesystem


Disk /dev/zram0: 8 GiB, 8589934592 bytes, 2097152 sectors
Units: sectors of 1 * 4096 = 4096 bytes
Sector size (logical/physical): 4096 bytes / 4096 bytes
I/O size (minimum/optimal): 4096 bytes / 4096 bytes

EDIT: Yes:

vincent@fedora:~$ systemd-analyze critical-chain
The time when unit became active or started is printed after the "@" character.
The time the unit took to start is printed after the "+" character.

graphical.target @1min 5.219s
└─udisks2.service @1min 5.222s +116ms
  └─basic.target @1.621s
    └─dbus-broker.service @1.603s +16ms
      └─dbus.socket @1.597s
        └─sysinit.target @1.594s
          └─systemd-resolved.service @1.552s +41ms
            └─run-credentials-systemd\x2dresolved.service.mount @2.407s

Don’t know why it did not popup at first time

EDIT 2: And graphical.target depends on it

vincent@fedora:~$ systemctl list-dependencies --reverse udisks2.service
udisks2.service
● └─graphical.target

After a long time, I found the solution, this line in dmesg:

[   62.533401] nvme nvme0: I/O tag 20 (2014) QID 0 timeout, completion polled

Meant in my case that Intel VMD was the issue. After disabling it in the BIOS (add to enter advanced options with CTRL+S), Fedora was starting fine and fast.

Windows broke on my dual boot, of course, but I followed this tutorial How to remove Intel VMD without reinstalling Windows | Scottie's Tech.Info (minus I had to suffer 3 restarts before getting startup options in recovery mode).

Now everything is fine.

1 Like