You are not logged in.
- Topics: Active | Unanswered
#1 2026-04-03 19:27:57
- graysky
- Wiki Maintainer
- From: :wq
- Registered: 2008-12-01
- Posts: 10,731
- Website
Debugging excessive boot time as a function of kernel space
I'm wondering if anyone has some suggestions for debugging what might be driving the 7 to 8 second kernel time that I'm experiencing on my workstation. For context, other machines I have running Arch have a kernel time on the order of several hundred milliseconds.
% systemd-analyze time
Startup finished in 21.531s (firmware) + 3.591s (loader) + 8.293s (kernel) + 3.133s (initrd) + 2.702s (userspace) = 39.252s
graphical.target reached after 2.701s in userspace.If I see any smoking guns, it would appear to be the R8169 interface, the NVMe drive, and the hrtimer. I don't quite know how to continue to dig in from here.
% dmesg --time-format=delta | grep -C 3 "+[1-8]\."
[ +0.002628] [< 0.002628>] AMD-Vi: Extended features (0x246577efa2254afa, 0x0): PPR NX GT [5] IA GA PC GA_vAPIC
[ +0.000006] [< 0.000006>] AMD-Vi: Interrupt remapping enabled
[ +0.076519] [< 0.076519>] Freeing initrd memory: 205112K
[ +4.062153] [< 4.062153>] hrtimer: interrupt took 1760 ns
[ +0.766974] [< 0.766974>] AMD-Vi: Virtual APIC enabled
[ +0.000002] [< 0.000002>] PCI-DMA: Using software bounce buffering for IO (SWIOTLB)
[ +0.000003] [< 0.000003>] software IO TLB: mapped [mem 0x0000000067773000-0x000000006b773000] (64MB)
--
[ +0.000061] [< 0.000061>] Loaded X.509 cert 'wens: 62a038651abadcf94bd0ac71f06c7248db18c200'
[ +0.030304] [< 0.030304>] Realtek Internal NBASE-T PHY r8169-0-700:00: attached PHY driver (mii_bus:phy_addr=r8169-0-700:00, irq=MAC)
[ +0.159187] [< 0.159187>] r8169 0000:07:00.0 enp7s0: Link is Down
[ +4.061995] [< 4.061995>] r8169 0000:07:00.0 enp7s0: Link is Up - 2.5Gbps/Full - flow control rx/tx
[ +0.740742] [< 0.740742>] usbcore: registered new interface driver snd-usb-audio
[ +4.214344] [< 4.214344>] nvme nvme0: using unchecked data bufferOffline