ACPI,SMP(?), and 5.2.1-RELEASE
David E. Cross
crossd at cs.rpi.edu
Thu Apr 8 19:21:30 PDT 2004
On Thu, 8 Apr 2004, Nate Lawson wrote:
> On Thu, 8 Apr 2004, David E. Cross wrote:
> Send the output of vmstat -i so I can see the interrupt load.
interrupt total rate
irq6: fdc0 9 0
irq8: rtc 20328 127
irq14: ata0 39 0
irq15: ata1 64 0
irq16: ahc1 407 2
irq17: pcm0 2 0
irq19: xl0 uhci0+ 2373 14
irq20: acpi0 6979146 43893
irq0: clk 15881 99
Total 7018239 44139
(ahc0 isn't shown, but that is on irq19 also, I am assuming that is what
the + is for)
> Try booting with:
> hint.apic.0.disabled="1"
> This is likely an APIC issue if this fixes it.
>
Yup, that fixes it.
> This also indicates an APIC problem. jhb@ has a patch that hopefully will
> be committed to allow disabling irq src overrides. Just for kicks, also
> reply with the output of "acpidump -t" on your system.
This is the result of running "acpidump -t" on a system with the APIC
disabled:
(this is typed by hand unfortunately)
/*
RSD PTR: OEM=ASUS, ACPI_Rev=1.0x (0)
RSDT=0x1fffd000, cksum=119
*/
/*
RSDT: Length=48, Revision=1, Checksum=182,
OEMID=ASUS, OEM Table ID=P2B-DS, OEM Revision=0x58582e32,
Creator ID=ASUS, Creator Revision=0x31303030
Entries={ 0x1fffd100, 0x1fffd040, 0x1fffd080 }
*/
/*
FADT: FACS=0x1ffff000, DSDT=0x1fffd180
INT_MODEL=PIC
Preferred_PM_Profile=Unspecified (0)
SCI_INT=9
SMI_CMD=0xb2, ACPI_ENABLE=0xa1, ACPI_DISABLE=0xa0, S4BIOS_REQ=0x0
PSTATE_CNT=0x0
PM1a_EVT_BLK=0xe400-0xe403
PM1a_CNT_BLK=0xe404-0xe405
PM_TMR_BLK=0xe408-0xe40b
GPE0_BLK=0xe40c-0xe40f
P_LVL2_LAT=190 us, P_LVL3_LAT=1900 us
FLUSH_SIZE=0, FLUSH_STRIDE=0
DUTY_OFFSET=1, DUTY_WIDTH=0
DAY_ALRM=13, MON_ALRM=0, CENTURY=0
IAPC_BOOT_ARCH=
Flags={WBINVD,PROC_C1,SLP_BUTTON,RTC_S4}
*/
/*
FACS: Length=64, HwSig=0x00000000, Firm_Wake_Vec=0x00000000
Global_Lock=
Flags=
Version=0
*/
/*
DSDT: Length=7645, Revision=1, Checksum=71,
OEMID=ASUS, OEM Table ID=P2B-DS, OEM Revision=0x1000,
Creator ID=MSFT, Creator Revision=0x1000001
*/
/*
BOOT: Length=40, Revision=1, Checksum=12,
OEMID=ASUS, OEM Table ID=P2B-DS, OEM Revision=0x58582e32,
Creator ID=ASUS, Creator Revision=0x31303030
*/
/*
APIC: Length=92, Revision=1, Checksum=58,
OEMID=ASUS, OEM Table ID=P2B-DS, OEM Revision=0x0,
Creator ID=, Creator Revision=0x0
Local APIC ADDR=0xfee00000
Flags={PC-AT}
Type=IO APIC
APIC ID=2
INT BASE=0
ADDR=0x00000000fec00000
Type=INT Override
BUS=0
IRQ=0
INTR=2
Flags={Polarity=conforming, Trigger=edge}
Type=INT Override
BUS=0
IRQ=9
INTR=20
Flags={Polarity=active-hi, Trigger=level}
Type=Local APIC
ACPI CPU=0
Flags={ENABLED}
APIC ID=1
Type=Local APIC
ACPI CPU=1
Flags={ENABLED}
APIC ID=0
*/
A brief glance suggests that this info is the same regardless of the
hint.apic.0.disabled flag. Also... hmm "Polartity=active-hi, and level
triggered" while on IRQ 20 would certainly explain an interrupt storm,
unless some other settings are changed.
Is there (was there) any possibility of data-loss because of this. Things
look fine, but is there the possibility that something went boom in the
background? restoring from backups is "easy" now, but if I continue to do
work on and there is something wrong in the background then it could get
painful in the future.
Is there anything wrong with running with APIC disabled? Should I run
with APIC disabled or ACPI disabled?
Thank you very much for your time.
--
David E. Cross
More information about the freebsd-acpi
mailing list