CONFIG_X86_PM_TIMER broken in 2.6.15.1 (since before 2.6.14?)
Hi all.
Now I have a repeatable confirmation that selecting CONFIG_X86_PM_TIMER=y does "break" printk on a few systems.
With CONFIG_X86_PM_TIMER=y pmtmr is preferred to tsc and as a result, printk time is not reset on CPU init.
The difference (hand diff) between the dmesg in the two configs is:
normal( # CONFIG_X86_PM_TIMER is not set ):
...
...
[17179569.184000] PID hash table entries: 4096 (order: 12, 65536 bytes)
[ 0.000000] Detected 3011.142 MHz processor.
[ 25.672059] Using tsc for high-res timesource
[ 25.673931] Console: colour VGA+ 80x25broken ( CONFIG_X86_PM_TIMER=y ):
...
[17179569.184000] ACPI: PM-Timer IO Port: 0x808 ... [17179569.184000] PID hash table entries: 4096 (order: 12, 65536 bytes)
[17179569.184000] Detected 3011.098 MHz processor.
[17179569.184000] Using pmtmr for high-res timesource
[17179569.184000] Console: colour VGA+ 80x25
Apart from that everything is the same (apart from the prink time).
Any clues as to why might this be broken?
I tired to look under arch/i386/kernel/timers but it was too much for me, got buried under too many jiffies :-)
Kalin.
Signature · 3 lines
|[ ~~~~~~~~~~~~~~~~~~~~~~ ]| +-> http://ThinRope.net/ <-+ |[ ______________________ ]|
Quoted 20 lines · Kalin KOZHUHAROV
Hi all.
Now I have a repeatable confirmation that selecting
CONFIG_X86_PM_TIMER=ydoes "break" printk on a few systems.With
CONFIG_X86_PM_TIMER=ypmtmr is preferred to tsc and as a result, printk time is not reset on CPU init.The difference (hand diff) between the dmesg in the two configs is:
normal( # CONFIG_X86_PM_TIMER is not set ):
........
[17179569.184000] PID hash table entries: 4096 (order: 12, 65536 bytes) [ 0.000000] Detected 3011.142 MHz processor. [ 25.672059] Using tsc for high-res timesource [ 25.673931] Console: colour VGA+ 80x25
Quoted 8 lines
broken (
CONFIG_X86_PM_TIMER=y):
....[17179569.184000] ACPI: PM-Timer IO Port: 0x808 .... [17179569.184000] PID hash table entries: 4096 (order: 12, 65536 bytes) [17179569.184000] Detected 3011.098 MHz processor. [17179569.184000] Using pmtmr for high-res timesource [17179569.184000] Console: colour VGA+ 80x25
My MUA broke the whitespace somehow, fixed above.
Quoted 6 lines
Apart from that everything is the same (apart from the prink time).
Any clues as to why might this be broken?
I tired to look under
arch/i386/kernel/timersbut it was too much for me, got buried under too many jiffies :-)
Now there is another system, my PIII laptop and it has another strange timer set:
ss ~ # dmesg |head -n 40
[17179569.184000] Linux version 2.6.15.1-K01_PIII_laptop (kalin@char) (gcc version 3.4.4 (Gentoo 3.4.4-r1, ssp-3.4.4-1.0, pie-8.7.8)) #1 PREEMPT Mon Jan 23 19:10:41 JST 2006 [17179569.184000] BIOS-provided physical RAM map: [17179569.184000] BIOS-e820: 0000000000000000 - 000000000009fc00 (usable) [17179569.184000] BIOS-e820: 000000000009fc00 - 00000000000a0000 (reserved) [17179569.184000] BIOS-e820: 00000000000e0000 - 00000000000eee00 (reserved) [17179569.184000] BIOS-e820: 00000000000eee00 - 00000000000ef000 (ACPI NVS) [17179569.184000] BIOS-e820: 00000000000ef000 - 0000000000100000 (reserved) [17179569.184000] BIOS-e820: 0000000000100000 - 000000000ef60000 (usable) [17179569.184000] BIOS-e820: 000000000ef60000 - 000000000ef70000 (ACPI data) [17179569.184000] BIOS-e820: 000000000ef70000 - 0000000010000000 (reserved) [17179569.184000] BIOS-e820: 00000000fff80000 - 0000000100000000 (reserved) [17179569.184000] 239MB LOWMEM available. [17179569.184000] On node 0 totalpages: 61280 [17179569.184000] DMA zone: 4096 pages, LIFO batch:0 [17179569.184000] DMA32 zone: 0 pages, LIFO batch:0 [17179569.184000] Normal zone: 57184 pages, LIFO batch:15 [17179569.184000] HighMem zone: 0 pages, LIFO batch:0 [17179569.184000] DMI 2.3 present. [17179569.184000] ACPI: RSDP (v000 TOSHIB ) @ 0x000f0090 [17179569.184000] ACPI: RSDT (v001 TOSHIB 750 0x00970814 TASM 0x04010000) @ 0x0ef60000 [17179569.184000] ACPI: FADT (v002 TOSHIB 750 0x00970814 TASM 0x04010000) @ 0x0ef60058 [17179569.184000] ACPI: DBGP (v001 TOSHIB 750 0x00970814 TASM 0x04010000) @ 0x0ef600dc [17179569.184000] ACPI: BOOT (v001 TOSHIB 750 0x00970814 TASM 0x04010000) @ 0x0ef60030 [17179569.184000] ACPI: DSDT (v001 TOSHIB 2000M 0x20020530 MSFT 0x0100000a) @ 0x00000000 [17179569.184000] Allocating PCI resources starting at 20000000 (gap: 10000000:eff80000) [17179569.184000] Built 1 zonelists [17179569.184000] Kernel command line: root=/dev/hda3 [17179569.184000] Local APIC disabled by BIOS -- you can enable it with "lapic" [17179569.184000] mapped APIC to ffffd000 (011e0000) [17179569.184000] Initializing CPU#0 [17179569.184000] PID hash table entries: 1024 (order: 10, 16384 bytes) [ 0.000000] Detected 796.776 MHz processor. [ 1364.722869] Using tsc for high-res timesource [ 1364.725059] Console: colour VGA+ 80x25 [ 1364.726889] Dentry cache hash table entries: 32768 (order: 5, 131072 bytes) [ 1364.727619] Inode-cache hash table entries: 16384 (order: 4, 65536 bytes) [ 1364.760246] Memory: 239904k/245120k available (1548k kernel code, 4792k reserved, 646k data, 132k init, 0k highmem) [ 1364.760361] Checking if this processor honours the WP bit even in supervisor mode... Ok. [ 1364.838652] Calibrating delay using timer specific routine.. 1596.68 BogoMIPS (lpj=3193360) [ 1364.838901] Security Framework v1.0.0 initialized
And, no it does not take 1364 seconds to boot, not that slow (about 35s?).
Is tsc totally broken?
I tried to fix the prink time by disabling the PM_TIMER but it didn't work that well.
ss ~ # gzcat /proc/config.gz |grep TIMECONFIG_HPET_TIMER=y
# CONFIG_X86_PM_TIMER is not setCONFIG_HANGCHECK_TIMER=yCONFIG_SND_TIMER=mCONFIG_SND_RTCTIMER=mCONFIG_SND_SEQ_RTCTIMER_DEFAULT=yCONFIG_PRINTK_TIME=y
Full .config available:
http://linux.tar.bz/reports/misc/ss/2.6.15.1-K01_PIII_laptop.config
other (lots of) stuff for this machine is in the directory:
http://linux.tar.bz/reports/misc/ss/
Kalin.
Signature · 3 lines
|[ ~~~~~~~~~~~~~~~~~~~~~~ ]| +-> http://ThinRope.net/ <-+ |[ ______________________ ]|