linux-kernel - Questions to TSC and clocksource and reducing boot time

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [day] [month] [year] [list]

Date:   Wed, 4 Apr 2018 17:20:54 +0200
From:   Paul Menzel <pmenzel+linux-x86@...gen.mpg.de>
To:     Thomas Gleixner <tglx@...utronix.de>,
        Ingo Molnar <mingo@...hat.com>,
        "H. Peter Anvin" <hpa@...or.com>
Cc:     x86@...nel.org, linux-kernel@...r.kernel.org
Subject: Questions to TSC and clocksource and reducing boot time

Dear Linux folks,


On the Lenovo X60 with Linux 4.16+ I try to reduce the boot time without 
having to rebuild the Linux kernel.

```
[    0.000000] tsc: Fast TSC calibration using PIT
[…]
[    0.000000] clocksource: hpet: mask: 0xffffffff max_cycles: 
0xffffffff, max_idle_ns: 133484882848 ns
[    0.000000] hpet clockevent registered
[    0.000000] tsc: Fast TSC calibration using PIT
[    0.004000] tsc: Detected 1662.367 MHz processor
[    0.004000] Calibrating delay loop (skipped), value calculated using 
timer frequency.. 3324.73 BogoMIPS (lpj=6649468)
[…]
[    0.153225] calling  init_acpi_pm_clocksource+0x0/0x172 @ 1
[    0.187744] clocksource: acpi_pm: mask: 0xffffff max_cycles: 
0xffffff, max_idle_ns: 2085701024 ns
[    0.187834] initcall init_acpi_pm_clocksource+0x0/0x172 returned 0 
after 33793 usecs
[…]
[    0.354787] tsc: Marking TSC unstable due to TSC halts in idle
[…]
[    1.376157] clocksource: tsc: mask: 0xffffffffffffffff max_cycles: 
0x17f648b8858, max_idle_ns: 440795225116 ns
```

Is there a way to disable the clocksource *tsc* as I know, that I’ll 
always get these results on the Lenovo X60.

Additionally, can I delay probing of the clocksource *acpi_pm*, so that 
the code is run at a later time, and possible 33 ms are saved from the 
critical/hot path?

Lastly, out of curiosity, why is the message `tsc: Fast TSC calibration 
using PIT` twice in the logs?


Kind regards,

Paul

View attachment "20180404–lenovo-x60–dmesg.txt" of type "text/plain" (131396 bytes)

Download attachment "smime.p7s" of type "application/pkcs7-signature" (5174 bytes)