Arch Linux

apaz

Member

Registered: 2018-07-23

Posts: 130

[SOLVED] Curiosity about Linux-firmware

Not important, but since the Linux-firmware split, the firmware loading at boot has increased by 4 sec (15s --> 19s). I would have thought it would have decreased! Does this happen to you too?

$ systemd-analyze
Startup finished in 18.879s (firmware) + 1.621s (loader) + 2.167s (kernel) + 2.087s (userspace) = 24.756s
graphical.target reached after 1.987s in userspace

Last edited by apaz (2022-02-16 09:22:12)

seth

Member

From: Won't reply 2 private help req

Registered: 2012-09-03

Posts: 77,063

Re: [SOLVED] Curiosity about Linux-firmware

That's not the linux-firmware but the firmware of your device, ie. the UEFI (then the bootloader in 1.6, the kernel in 2.1 and the rest within 2 seconds.

---

How to upload text · How to boot w/o GUI · Disable Windows Fast-Start! · Fix your xinitrc

Slithery

Administrator

From: Norfolk, UK

Registered: 2013-12-01

Posts: 5,776

Re: [SOLVED] Curiosity about Linux-firmware

The firmware time reported by systemd-analyze has absolutely nothing to do with the linux-firmware package.
It's the time that it takes for your UEFI to initialize your system and hand off to the bootloader.

---

No, it didn't "fix" anything. It just shifted the brokeness one space to the right. - jasonwryan
Closing -- for deletion; Banning -- for muppetry. - jasonwryan

aur - dotfiles

apaz

Member

Registered: 2018-07-23

Posts: 130

Re: [SOLVED] Curiosity about Linux-firmware

OK, thanks. So Linux-firmware is part of the "kernel" step? But the increase in boot time coincides with the Linux-firmware split and I haven't updated the bios in months. Is it possible that there is a relationship?

EDIT
I don't know if it helps, but I'll add the result of dmesg /firmware:

$ sudo dmesg | grep -i firmware
[sudo] password di paz:
[    0.459446] ACPI: [Firmware Bug]: BIOS _OSI(Linux) query ignored
[    0.466471] acpi PNP0A08:00: [Firmware Info]: MMCONFIG for domain 0000 [bus 00-7f] only partially covers this bridge
[    1.246760] [drm] Found VCN firmware Version ENC: 1.14 DEC: 5 VEP: 0 Revision: 20
[    1.246765] amdgpu 0000:0c:00.0: amdgpu: Will use PSP to load VCN firmware
[    2.898186] iwlwifi 0000:05:00.0: loaded firmware version 46.4e1ceb39.0 9260-th-b0-jf-b0-46.ucode op_mode iwlmvm
[    3.077642] Bluetooth: hci0: Found device firmware: intel/ibt-18-16-1.sfi
[    3.077673] Bluetooth: hci0: Firmware Version: 86-46.21
[    3.077674] Bluetooth: hci0: Firmware already loaded

Last edited by apaz (2022-02-16 09:18:42)

V1del

Forum Moderator

Registered: 2012-10-16

Posts: 25,292

Re: [SOLVED] Curiosity about Linux-firmware

The firmware files got compressed and need to be uncompressed on load, maybe you are seeing that, but check/compare actual dmesg/kernel logs to actually know. Also very likely that the kernel in general changed if you had a 5.15 -> 5.16 jump you'll almost certainly see an effect, just by virtue of having changes in the kernel.

The only way you could clearly make any inference is by reverting your kernel to the exact same version that was "faster" and compare then.

But again, if it's just the "firmware" loading time that has increased then this is something in your UEFI, could also just be the difference between a cold and a warm boot, did you add any devices? Have a certain USB plugged in you didn't have with the lower time?

Last edited by V1del (2022-02-16 09:18:26)

apaz

Member

Registered: 2018-07-23

Posts: 130

Re: [SOLVED] Curiosity about Linux-firmware

You're right: with the introduction of linux-firmware split there was also a switch between kernel 5.15 and 5.16. So it depends on that...
Thanks, I consider my curiosity satisfied and put a [Solved].