© 2004
Canonical Ltd.
•
Terms of use
•
Data privacy
•
Contact Launchpad Support
•
Blog
•
Careers
•
System status
•
0e1f616
(Get the code!)
Sorry, but the bug also happens with 4.15.0-23
I upgraded a PC with a SSD replacing a HDD. So this laptop now only has a SSD which obviously has no spinning part to produce entropy to seed the random generator.
Here is what happens:
-1) not touching anything, there is a "30 seconds gap" (visible in dmesg) where nothing at all happens, then the boot sequence continues
-2) touching the trackpad randomly during the boot sequence makes the random generator initialize quicker (message in dmesg) but there is still the 30 sec gap in spite of the message in the log
-3) plugging an external spinning rust to USB fixes the issue! Now the system does not hang for 30 sec waiting anymore...
DMESG with 1)
[ 0.000000] microcode: microcode updated early to revision 0x44, date = 2010-10-04
[ 0.000000] Linux version 4.15.0-23-generic (buildd@
[ 0.000000] Command line: BOOT_IMAGE=
[ 0.000000] KERNEL supported cpus:
[ 0.000000] Intel GenuineIntel
[ 0.000000] AMD AuthenticAMD
[ 0.000000] Centaur CentaurHauls
[ 0.000000] x86/fpu: x87 FPU will use FXSAVE
(...)
[ 2.742732] fbcon: inteldrmfb (fb0) is primary device
[ 2.743323] Console: switching to colour frame buffer device 160x50
[ 2.743365] i915 0000:00:02.0: fb0: inteldrmfb frame buffer device
[ 2.816170] firewire_core 0000:02:09.0: created device fw0: GUID 474fc0001ac27070, S400
[ 3.421641] input: AlpsPS/2 ALPS GlidePoint as /devices/
[ 34.943967] EXT4-fs (sda1): mounted filesystem with ordered data mode. Opts: (null)
[ 35.129322] ip_tables: (C) 2000-2006 Netfilter Core Team
[ 35.142828] systemd[1]: systemd 237 running in system mode. (+PAM +AUDIT +SELINUX +IMA +APPARMOR +SMACK +SYSVINIT +UTMP +LIBCRYPTSETUP +GCRYPT +GNUTLS +ACL +XZ +LZ4 +SECCOMP +BLKID +ELFUTILS +KMOD -IDN2 +IDN -PCRE2 default-
[ 35.160097] systemd[1]: Detected architecture x86-64.
(...)
[ 46.267417] sky2 0000:09:00.0 enp9s0: Link is up at 100 Mbps, full duplex, flow control both
[ 46.267444] IPv6: ADDRCONF(
DMESG with 2)
[ 0.000000] microcode: microcode updated early to revision 0x44, date = 2010-10-04
[ 0.000000] Linux version 4.15.0-23-generic (buildd@
[ 0.000000] Command line: BOOT_IMAGE=
[ 0.000000] KERNEL supported cpus:
[ 0.000000] Intel GenuineIntel
[ 0.000000] AMD AuthenticAMD
[ 0.000000] Centaur CentaurHauls
[ 0.000000] x86/fpu: x87 FPU will use FXSAVE
(...)
[ 2.749189] fbcon: inteldrmfb (fb0) is primary device
[ 2.749779] Console: switching to colour frame buffer device 160x50
[ 2.749822] i915 0000:00:02.0: fb0: inteldrmfb frame buffer device
[ 2.848172] firewire_core 0000:02:09.0: created device fw0: GUID 474fc0001ac27070, S400
[ 3.510278] input: AlpsPS/2 ALPS GlidePoint as /devices/
[ 6.739410] random: crng init done
[ 34.948181] EXT4-fs (sda1): mounted filesystem with ordered data mode. Opts: (null)
[ 35.133775] ip_tables: (C) 2000-2006 Netfilter Core Team
[ 35.147429] systemd[1]: systemd 237 running in system mode. (+PAM +AUDIT +SELINUX +IMA +APPARMOR +SMACK +SYSVINIT +UTMP +LIBCRYPTSETUP +GCRYPT +GNUTLS +ACL +XZ +LZ4 +SECCOMP +BLKID +ELFUTILS +KMOD -IDN2 +IDN -PCRE2 default-
[ 35.164102] systemd[1]: Detected architecture x86-64.
(...)
[ 46.364727] sky2 0000:09:00.0 enp9s0: Link is up at 100 Mbps, full duplex, flow control both
[ 46.364753] IPv6: ADDRCONF(
DMESG with 3)
[ 0.000000] microcode: microcode updated early to revision 0x44, date = 2010-10-04
[ 0.000000] Linux version 4.15.0-23-generic (buildd@
[ 0.000000] Command line: BOOT_IMAGE=
[ 0.000000] KERNEL supported cpus:
[ 0.000000] Intel GenuineIntel
[ 0.000000] AMD AuthenticAMD
[ 0.000000] Centaur CentaurHauls
[ 0.000000] x86/fpu: x87 FPU will use FXSAVE
(...)
[ 2.715313] fbcon: inteldrmfb (fb0) is primary device
[ 2.715908] Console: switching to colour frame buffer device 160x50
[ 2.715950] i915 0000:00:02.0: fb0: inteldrmfb frame buffer device
[ 2.784151] firewire_core 0000:02:09.0: created device fw0: GUID 474fc0001ac27070, S400
[ 2.869336] usb-storage 2-2:1.0: USB Mass Storage device detected
[ 2.869486] scsi host5: usb-storage 2-2:1.0
[ 2.870357] usbcore: registered new interface driver usb-storage
[ 2.873260] usbcore: registered new interface driver uas
[ 3.394728] input: AlpsPS/2 ALPS GlidePoint as /devices/
[ 5.408440] scsi 5:0:0:0: Direct-Access Inateck 1A11 PQ: 0 ANSI: 6
[ 5.408861] sd 5:0:0:0: Attached scsi generic sg2 type 0
[ 5.409416] sd 5:0:0:0: [sdb] 156301488 512-byte logical blocks: (80.0 GB/74.5 GiB)
[ 5.410425] sd 5:0:0:0: [sdb] Write Protect is off
[ 5.410430] sd 5:0:0:0: [sdb] Mode Sense: 43 00 00 00
[ 5.411424] sd 5:0:0:0: [sdb] Write cache: disabled, read cache: enabled, doesn't support DPO or FUA
[ 7.008662] sdb: sdb1 sdb2 < sdb5 sdb6 sdb7 sdb8 sdb9 >
[ 7.012088] sd 5:0:0:0: [sdb] Attached SCSI disk
[ 8.200970] EXT4-fs (sda1): mounted filesystem with ordered data mode. Opts: (null)
[ 8.383076] ip_tables: (C) 2000-2006 Netfilter Core Team
[ 8.396816] systemd[1]: systemd 237 running in system mode. (+PAM +AUDIT +SELINUX +IMA +APPARMOR +SMACK +SYSVINIT +UTMP +LIBCRYPTSETUP +GCRYPT +GNUTLS +ACL +XZ +LZ4 +SECCOMP +BLKID +ELFUTILS +KMOD -IDN2 +IDN -PCRE2 default-
[ 8.416102] systemd[1]: Detected architecture x86-64.
(...)
[ 19.753437] iwl3945 0000:0b:00.0 wlp11s0: renamed from wlan0
[ 19.981818] sky2 0000:09:00.0 enp9s0: Link is up at 100 Mbps, full duplex, flow control both
-------
So, as you can see... unfortunately changing from a HDD to a SSD does NOT save time due to this nasty bug!
Also, for kernel maintainers, this is probably not obvious to spot since it shows more blatantly on a machine that has NO HDD at all. Usually I would assume a developer's machine (as mine) always has a SDD for big files that don't change often, and SSD for speed. Since you have a HDD (even unused for boot sequence) the problem don't show.
Indeed, the fact to simply plug a spinning device to USB fixes the bug (note that plugging a "non-spinning" USB key does not fix anything, as expected!)
The 3 dmesg shown above prove that the "fix" (plugging a spinning disk) reduces the boot sequence from 46 sec to 20 sec, saving around 25 seconds!
So I hope kernel gurus will fix that soon so that I can enjoy the full speed of my brand new SSD!