linux-kernel - Re: C1E auto-promotion suspend/resume

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

Date:	Sun, 13 Mar 2016 22:24:47 -0700
From:	Andy Lutomirski <luto@...capital.net>
To:	"Brown, Len" <len.brown@...el.com>
Cc:	"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
Subject: Re: C1E auto-promotion suspend/resume

On Sun, Mar 13, 2016 at 9:31 PM, Brown, Len <len.brown@...el.com> wrote:
>> By BIOS (1.2.3 on a Dell XPS 13 9350) seems to want to enable C1E
>> auto-promotion (ugh!), which results in this difference across
>> suspend/resume according to turbostat:
>>
>> -cpu3: MSR_IA32_POWER_CTL: 0x0024005d (C1E auto-promotion: DISabled)
>> +cpu3: MSR_IA32_POWER_CTL: 0x0024005f (C1E auto-promotion: ENabled)
>>
>> Should intel_idle learn to re-disable idle promotion on resume?
>
> Yes, it seems that way.
>
> Go ahead and send a patch, or file a bug at bugzilla.kernel.org
> and we'll get to it.

Sent.  The only other differences I see across suspend/resume in
turbostat --debug (with taskset -c 0 to suppress spurious junk) are:

--- pre-susp.txt    2016-03-13 22:21:39.889337697 -0700
+++ post-susp.txt    2016-03-13 21:38:20.782503438 -0700
@@ -24,8 +24,8 @@
 cpu0: MSR_NHM_SNB_PKG_CST_CFG_CTL: 0x1e008006 (UNdemote-C3,
UNdemote-C1, demote-C3, demote-C1, locked: pkg-cstate-limit=6: pc8)
 cpu0: MSR_PM_ENABLE: 0x00000001 (HWP)
 cpu0: MSR_HWP_CAPABILITIES: 0x0108171c (high 0x1c guar 0x17 eff 0x8 low 0x1)
-cpu0: MSR_HWP_REQUEST: 0x80001c04 (min 0x4 max 0x1c des 0x0 epp 0x80
window 0x0 pkg 0x0)
-cpu0: MSR_HWP_INTERRUPT: 0x00000001 (EN_Guaranteed_Perf_Change,
Dis_Excursion_Min)
+cpu0: MSR_HWP_REQUEST: 0x8000ff01 (min 0x1 max 0xff des 0x0 epp 0x80
window 0x0 pkg 0x0)
+cpu0: MSR_HWP_INTERRUPT: 0x00000000 (Dis_Guaranteed_Perf_Change,
Dis_Excursion_Min)
 cpu0: MSR_HWP_STATUS: 0x00000000 (No-Guaranteed_Perf_Change, No-Excursion_Min)
 cpu0: MSR_IA32_ENERGY_PERF_BIAS: 0x00000006 (balanced)
 cpu0: MSR_RAPL_POWER_UNIT: 0x000a0e03 (0.125000 Watts, 0.000061
Joules, 0.000977 sec.)
@@ -36,6 +36,6 @@
 cpu0: MSR_DRAM_POWER_LIMIT: 0x5400de00000000 (UNlocked)
 cpu0: DRAM Limit: DISabled (0.000000 Watts, 0.000977 sec, clamp DISabled)
 cpu0: MSR_IA32_TEMPERATURE_TARGET: 0x00640000 (100 C)
-cpu0: MSR_IA32_PACKAGE_THERM_STATUS: 0x88360000 (46 C)
-cpu0: MSR_IA32_THERM_STATUS: 0x88360000 (46 C +/- 1)
-cpu1: MSR_IA32_THERM_STATUS: 0x88360000 (46 C +/- 1)
+cpu0: MSR_IA32_PACKAGE_THERM_STATUS: 0x883c0000 (40 C)
+cpu0: MSR_IA32_THERM_STATUS: 0x883c0000 (40 C +/- 1)
+cpu1: MSR_IA32_THERM_STATUS: 0x88400000 (36 C +/- 1)


Are either of those potentially interesting?