Re: Boot panic on MP amd64 with 5.5, snapshot kernels
On Sun, May 18, 2014 at 03:04:30PM -0400, John D. Verne wrote: I just got a new amd64 box to run OpenBSD on, but it is panicking on boot when I try to run the 5.5 kernel on it. The panic is unknown MPS interrupt trigger 2 somewhere in the acpi code. I see the same panic on this Asus J1800I-C. http://www.asus.com/Motherboards/J1800IC/specifications/ http://marc.info/?t=14042978995r=1w=2 NetBSD 6.1.4 manages to enumerate all the ACPI stuff Yes, it booted for me too; here is the acpidump -dt http://stare.cz/dmesg/asus-J1800IC-asl.gz https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=187966 Yes, this seems to be the bug. My ACPI table says: Type=Local APIC NMI ACPI CPU=1 LINT Pin=72 Flags={Polarity=active-hi, Trigger=level} Type=Local APIC NMI ACPI CPU=2 LINT Pin=55 Flags={Polarity=0x2, Trigger=0x2} Not that I understand what polarity and trigger are, but if I enable just 1 core in the BIOS, this panic disappears. The second CPU's polarity and trigger values seem to be what the panic is complaining about. I've also booted the OpenBSD snapshot from May 19 by disabling the acpi0 device via UKC, On this board, disabling acpi make the kernel panic in identifycpu(). and then tweaked the kernel in the same manner FreeBSD does, which allows the boot process to not panic with acpi enabled. So, copying what Linux and FreeBSD does naively fixes things. I'll leave the rest up to the experts. Blindly using John's changes described in http://marc.info/?l=openbsd-techm=140115427622786w=2 also makes my kernel not panic on the ACPI tables. However, then I ran into another panic related to lapic. During the FreeBSD-current back-and-forth, I ended up disabling half the serial ports on this motherboard via the BIOS. It looks like the three back panel serial ports are acceptable, but the three on-board serial ports cause a panic. FreeBSD hangs when enumerating those, and OpenBSD panics. I'll raise this as a seperate issue, but for now I've disabled them. Getting the same lapic panic, I also tried disabling the serial ports, but it didn't. With John's lapic change (printf a warning instead) the machine boots, with the serial ports enabled - dmesg below. Thanks! However the serial port and the USB ports do not work. Maybe here is the point in the boot sequence: lapic_set_lvt: bad pin value 72 ehci_sync_hc: tsleep() = 35 ehci_sync_hc: tsleep() = 35 ehci_sync_hc: tsleep() = 35 ehci_sync_hc: tsleep() = 35 ehci_sync_hc: tsleep() = 35 ehci_sync_hc: tsleep() = 35 ehci_sync_hc: tsleep() = 35 ehci_sync_hc: tsleep() = 35 uhub0: device problem, disabling port 1 I know for a fact that the USB ports work, as I can use a mouse in them (in the new fancy graphic BIOS). But in the booted OpenBSD, they donlt seem to be present. Anyway, thanks for making my machine boot! Jan OpenBSD 5.5-current (GENERIC.MP) #0: Tue Jul 8 23:12:34 CEST 2014 r...@media.stare.cz:/usr/src/sys/arch/i386/compile/GENERIC.MP cpu0: Intel(R) Celeron(R) CPU J1800 @ 2.41GHz (GenuineIntel 686-class) 2.42 GHz cpu0: FPU,V86,DE,PSE,TSC,MSR,PAE,MCE,CX8,APIC,SEP,MTRR,PGE,MCA,CMOV,PAT,PSE36,CFLUSH,DS,ACPI,MMX,FXSR,SSE,SSE2,SS,HTT,TM,PBE,NXE,LONG,SSE3,PCLMUL,DTES64,MWAIT,DS-CPL,VMX,EST,TM2,SSSE3,CX16,xTPR,PDCM,SSE4.1,SSE4.2,MOVBE,POPCNT,DEADLINE,RDRAND,LAHF,3DNOWP,PERF,ITSC,SMEP,ERMS real mem = 2002079744 (1909MB) avail mem = 1956917248 (1866MB) mpath0 at root scsibus0 at mpath0: 256 targets mainbus0 at root bios0 at mainbus0: AT/286+ BIOS, date 12/14/12, SMBIOS rev. 2.8 @ 0xebd60 (43 entries) bios0: vendor American Megatrends Inc. version 0604 date 06/10/2014 bios0: ASUS All Series acpi0 at bios0: rev 2 acpi0: sleep states S0 S3 S4 S5 acpi0: tables DSDT FACP APIC FPDT MCFG LPIT HPET SSDT SSDT SSDT UEFI acpi0: wakeup devices UAR5(S4) UAR8(S4) PS2K(S4) PS2M(S4) UAR1(S4) URIR(S4) XHC1(S4) EHC1(S4) PXSX(S4) PXSX(S4) PXSX(S4) PXSX(S4) PWRB(S0) acpitimer0 at acpi0: 3579545 Hz, 24 bits acpimadt0 at acpi0 addr 0xfee0: PC-AT compat cpu0 at mainbus0: apid 0 (boot processor) mtrr: Pentium Pro MTRR support, 8 var ranges, 88 fixed ranges cpu0: apic clock running at 83MHz cpu0: mwait min=64, max=64, C-substates=0.2.0.0.0, IBE cpu1 at mainbus0: apid 2 (application processor) cpu1: Intel(R) Celeron(R) CPU J1800 @ 2.41GHz (GenuineIntel 686-class) 2.42 GHz cpu1: FPU,V86,DE,PSE,TSC,MSR,PAE,MCE,CX8,APIC,SEP,MTRR,PGE,MCA,CMOV,PAT,PSE36,CFLUSH,DS,ACPI,MMX,FXSR,SSE,SSE2,SS,HTT,TM,PBE,NXE,LONG,SSE3,PCLMUL,DTES64,MWAIT,DS-CPL,VMX,EST,TM2,SSSE3,CX16,xTPR,PDCM,SSE4.1,SSE4.2,MOVBE,POPCNT,DEADLINE,RDRAND,LAHF,3DNOWP,PERF,ITSC,SMEP,ERMS ioapic0 at mainbus0: apid 1 pa 0xfec0, version 20, 87 pins reserved polarity 2, assuming low polarity reserved trigger 2, assuming level trigger acpimcfg0 at acpi0 addr 0xe000, bus 0-255 acpihpet0 at acpi0: 14318179 Hz acpiprt0 at acpi0: bus 0 (PCI0) acpiprt1 at acpi0: bus 1 (RP01) acpiprt2 at acpi0: bus 2 (RP02) acpiprt3
Re: Boot panic on MP amd64 with 5.5, snapshot kernels
NetBSD 6.1.4 manages to enumerate all the ACPI stuff Yes, it booted for me too; here is the acpidump -dt http://stare.cz/dmesg/asus-J1800IC-asl.gz the machine boots, dmesg below. Thanks! So now that I can boot the (tweaked) OpenBSD kernel: http://stare.cz/dmesg/asus-J1800IC-acpidump.tar.gz While disassembling DSDT, iasl complained: Loading Acpi table from file asus-J1800IC.DSDT.2 - Length 00034068 (008514) ACPI: DSDT 0x 008514 (v02 ALASKA A M I01072009 INTL 20120913) Acpi table [DSDT] successfully installed and loaded Pass 1 parse of [DSDT] Pass 2 parse of [DSDT] Parsing Deferred Opcodes (Methods/Buffers/Packages/Regions) Parsing completed Found 1 external control methods, reparsing with new information Pass 1 parse of [DSDT] Pass 2 parse of [DSDT] Parsing Deferred Opcodes (Methods/Buffers/Packages/Regions) Parsing completed Disassembly completed ASL Output:asus-J1800IC.DSDT.dsl - 353462 bytes iASL Warning: There were 1 external control methods found during disassembly, but additional ACPI tables to resolve these externals were not specified. The resulting disassembler output file may not compile because the disassembler did not know how many arguments to assign to these methods. To specify the tables needed to resolve external control method references, the -e option can be used to specify the filenames. Example iASL invocations: iasl -e ssdt1.aml ssdt2.aml ssdt3.aml -d dsdt.aml iasl -e dsdt.aml ssdt2.aml -d ssdt1.aml iasl -e ssdt*.aml -d dsdt.aml In addition, the -fe option can be used to specify a file containing control method external declarations with the associated method argument counts. Each line of the file must be of the form: External (method pathname, MethodObj, argument count) Invocation: iasl -fe refs.txt -d dsdt.aml
Re: Boot panic on MP amd64 with 5.5, snapshot kernels
On Tue, May 20, 2014 at 09:51:50AM -0400, John D. Verne wrote: On Mon, May 19, 2014 at 11:19:23AM -0700, Mike Larkin wrote: On Mon, May 19, 2014 at 01:42:49PM -0400, John D. Verne wrote: On Sun, May 18, 2014 at 03:04:30PM -0400, John D. Verne wrote: I just got a new amd64 box to run OpenBSD on, but it is panicking on boot when I try to run the 5.5 kernel on it. The panic is unknown MPS interrupt trigger 2 somewhere in the acpi code. bsd.rd, bsd and bsd.mp all panic in the same places, as does bsd.rd from the latest amd64 snapshot. NetBSD 6.1.4 manages to enumerate all the ACPI stuff when I use their boot image. So there's that. As an aside, I was surprised at how different the src/sys tree is from OpenBSD. But I'm going to try and see how they handle the Intel ACPICA 20110623 device, which seems to be the thing that is not working right. Get a dump of the AML using FreeBSD then. Can't really help otherwise. Ok, thanks for the tip. I managed to get FreeBSD 11-current to boot, and I've run acpidump -td to get the attached output. I hope this is what you wanted. If not, I have the FreeBSD liveCD running on the OpenBSD snapshot install on this hardware. I am at your disposal. I've also booted the OpenBSD snapshot from May 19 by disabling the acpi0 device via UKC, and then tweaked the kernel in the same manner FreeBSD does, which allows the boot process to not panic with acpi enabled. So, copying what Linux and FreeBSD does naively fixes things. I'll leave the rest up to the experts. However, then I ran into another panic related to lapic. During the FreeBSD-current back-and-forth, I ended up disabling half the serial ports on this motherboard via the BIOS. It looks like the three back panel serial ports are acceptable, but the three on-board serial ports cause a panic. FreeBSD hangs when enumerating those, and OpenBSD panics. I'll raise this as a seperate issue, but for now I've disabled them. Admittedly, this is a weird board, so those ports are both highly configurable and probably presented to (what looks like) the ISA bus in an odd manner. Perhaps an email of such a tremendous size is unlikely to get to the list. I'll try attaching the acpidump output. [demime 1.01d removed an attachment of type application/x-gunzip]
Re: Boot panic on MP amd64 with 5.5, snapshot kernels
On Tue, May 20, 2014 at 12:26:04PM -0400, John D. Verne wrote: On Tue, May 20, 2014 at 09:51:50AM -0400, John D. Verne wrote: On Mon, May 19, 2014 at 11:19:23AM -0700, Mike Larkin wrote: On Mon, May 19, 2014 at 01:42:49PM -0400, John D. Verne wrote: On Sun, May 18, 2014 at 03:04:30PM -0400, John D. Verne wrote: I just got a new amd64 box to run OpenBSD on, but it is panicking on boot when I try to run the 5.5 kernel on it. The panic is unknown MPS interrupt trigger 2 somewhere in the acpi code. bsd.rd, bsd and bsd.mp all panic in the same places, as does bsd.rd from the latest amd64 snapshot. NetBSD 6.1.4 manages to enumerate all the ACPI stuff when I use their boot image. So there's that. As an aside, I was surprised at how different the src/sys tree is from OpenBSD. But I'm going to try and see how they handle the Intel ACPICA 20110623 device, which seems to be the thing that is not working right. Get a dump of the AML using FreeBSD then. Can't really help otherwise. Ok, thanks for the tip. I managed to get FreeBSD 11-current to boot, and I've run acpidump -td to get the attached output. I hope this is what you wanted. If not, I have the FreeBSD liveCD running on the OpenBSD snapshot install on this hardware. I am at your disposal. I've also booted the OpenBSD snapshot from May 19 by disabling the acpi0 device via UKC, and then tweaked the kernel in the same manner FreeBSD does, which allows the boot process to not panic with acpi enabled. So, copying what Linux and FreeBSD does naively fixes things. I'll leave the rest up to the experts. However, then I ran into another panic related to lapic. During the FreeBSD-current back-and-forth, I ended up disabling half the serial ports on this motherboard via the BIOS. It looks like the three back panel serial ports are acceptable, but the three on-board serial ports cause a panic. FreeBSD hangs when enumerating those, and OpenBSD panics. I'll raise this as a seperate issue, but for now I've disabled them. Admittedly, this is a weird board, so those ports are both highly configurable and probably presented to (what looks like) the ISA bus in an odd manner. Perhaps an email of such a tremendous size is unlikely to get to the list. I'll try attaching the acpidump output. Perhaps MIME attachments are rejected. Here's a URL: http://www.clevermonkey.org/OpenBSD/ACPI_ASRock_IMB-150.txt.gz Sorry for the noise. -- John D. Verne j...@clevermonkey.org
Re: Boot panic on MP amd64 with 5.5, snapshot kernels
On Sun, May 18, 2014 at 03:04:30PM -0400, John D. Verne wrote: I just got a new amd64 box to run OpenBSD on, but it is panicking on boot when I try to run the 5.5 kernel on it. The panic is unknown MPS interrupt trigger 2 somewhere in the acpi code. bsd.rd, bsd and bsd.mp all panic in the same places, as does bsd.rd from the latest amd64 snapshot. NetBSD 6.1.4 manages to enumerate all the ACPI stuff when I use their boot image. So there's that. As an aside, I was surprised at how different the src/sys tree is from OpenBSD. But I'm going to try and see how they handle the Intel ACPICA 20110623 device, which seems to be the thing that is not working right. -- John D. Verne j...@clevermonkey.org
Re: Boot panic on MP amd64 with 5.5, snapshot kernels
On Mon, May 19, 2014 at 01:42:49PM -0400, John D. Verne wrote: On Sun, May 18, 2014 at 03:04:30PM -0400, John D. Verne wrote: I just got a new amd64 box to run OpenBSD on, but it is panicking on boot when I try to run the 5.5 kernel on it. The panic is unknown MPS interrupt trigger 2 somewhere in the acpi code. bsd.rd, bsd and bsd.mp all panic in the same places, as does bsd.rd from the latest amd64 snapshot. NetBSD 6.1.4 manages to enumerate all the ACPI stuff when I use their boot image. So there's that. As an aside, I was surprised at how different the src/sys tree is from OpenBSD. But I'm going to try and see how they handle the Intel ACPICA 20110623 device, which seems to be the thing that is not working right. Get a dump of the AML using FreeBSD then. Can't really help otherwise. And PS there is no such thing as an Intel ACPICA 20110623 device. That's the parser code NetBSD is using. We have our own. -- John D. Verne j...@clevermonkey.org
Re: Boot panic on MP amd64 with 5.5, snapshot kernels
On Mon, May 19, 2014 at 11:19:23AM -0700, Mike Larkin wrote: On Mon, May 19, 2014 at 01:42:49PM -0400, John D. Verne wrote: On Sun, May 18, 2014 at 03:04:30PM -0400, John D. Verne wrote: I just got a new amd64 box to run OpenBSD on, but it is panicking on boot when I try to run the 5.5 kernel on it. The panic is unknown MPS interrupt trigger 2 somewhere in the acpi code. bsd.rd, bsd and bsd.mp all panic in the same places, as does bsd.rd from the latest amd64 snapshot. NetBSD 6.1.4 manages to enumerate all the ACPI stuff when I use their boot image. So there's that. As an aside, I was surprised at how different the src/sys tree is from OpenBSD. But I'm going to try and see how they handle the Intel ACPICA 20110623 device, which seems to be the thing that is not working right. Get a dump of the AML using FreeBSD then. Can't really help otherwise. Well, the FreeBSD boot panics in a similar sort of way. Bogus interrupt trigger mode. This is from memory -- I didn't get a full copy of the boot messages, though it did have acpi and madt in backtrace. I can get it if this is important, though this is starting to look like a Windows-only box. -- John D. Verne j...@clevermonkey.org
Re: Boot panic on MP amd64 with 5.5, snapshot kernels
On Mon, May 19, 2014 at 03:56:05PM -0400, John D. Verne wrote: On Mon, May 19, 2014 at 11:19:23AM -0700, Mike Larkin wrote: On Mon, May 19, 2014 at 01:42:49PM -0400, John D. Verne wrote: On Sun, May 18, 2014 at 03:04:30PM -0400, John D. Verne wrote: I just got a new amd64 box to run OpenBSD on, but it is panicking on boot when I try to run the 5.5 kernel on it. The panic is unknown MPS interrupt trigger 2 somewhere in the acpi code. bsd.rd, bsd and bsd.mp all panic in the same places, as does bsd.rd from the latest amd64 snapshot. NetBSD 6.1.4 manages to enumerate all the ACPI stuff when I use their boot image. So there's that. As an aside, I was surprised at how different the src/sys tree is from OpenBSD. But I'm going to try and see how they handle the Intel ACPICA 20110623 device, which seems to be the thing that is not working right. Get a dump of the AML using FreeBSD then. Can't really help otherwise. Well, the FreeBSD boot panics in a similar sort of way. Bogus interrupt trigger mode. This is from memory -- I didn't get a full copy of the boot messages, though it did have acpi and madt in backtrace. I can get it if this is important, though this is starting to look like a Windows-only box. However, the FreeBSD message allowed my Google-fu to work. This looks apropos: http://forums.freenas.org/index.php?threads/kernel-panic-bogus-interrupt-trigger-mode-on-intel-j1900.20851/ And http://www.freebsd.org/cgi/query-pr.cgi?pr=187966 Nearly the exact hardware I have. Which also appears to have a buggy BIOS. -- John D. Verne j...@clevermonkey.org
Boot panic on MP amd64 with 5.5, snapshot kernels
I just got a new amd64 box to run OpenBSD on, but it is panicking on boot when I try to run the 5.5 kernel on it. The panic is unknown MPS interrupt trigger 2 somewhere in the acpi code. bsd.rd, bsd and bsd.mp all panic in the same places, as does bsd.rd from the latest amd64 snapshot. I haven't tried i386 yet. I messed around with boot -c and verbose mode, and captured the entire dmseg from ddb from 5.5 /bsd. I've placed the lame camera phone pics here: http://www.clevermonkey.org/OpenBSD/amd64_55_info/ I couldn't figure out how to save the actual text. Sorry. Also, if it matters, this is all done via PXE boot, as this box has no CDROM or floppy, and I didn't have the facility to make an amd64 USB stick (this is my only amd64 box.) Also, if it matters, I opted to not get the Atom version of this board because I wanted some of the more interesting chipset support. It is a Celeron J1000. I also wanted the better supported Intel graphics stuff (even though I suspect this box will run in text-only mode most of its life.) I'm staring at the ACPI code trying to figure out the various defines used, and 2 seems to be there. I welcome conversations about what this code is doing and what it wants out of total curiousity. -- John D. Verne j...@clevermonkey.org
Re: Boot panic on MP amd64 with 5.5, snapshot kernels
On Sun, May 18, 2014 at 03:04:30PM -0400, John D. Verne wrote: Also, if it matters, I opted to not get the Atom version of this board because I wanted some of the more interesting chipset support. It is a Celeron J1000. I also wanted the better supported Intel graphics stuff (even though I suspect this box will run in text-only mode most of its life.) Uh, sorry. J1900, the embedded version. -- John D. Verne j...@clevermonkey.org
Re: Boot panic on MP amd64 with 5.5, snapshot kernels
On Sun, May 18, 2014 at 03:04:30PM -0400, John D. Verne wrote: I'm staring at the ACPI code trying to figure out the various defines Uh, I actually meant APIC. I need more sleep. -- John D. Verne j...@clevermonkey.org
Re: Boot panic on MP amd64 with 5.5, snapshot kernels
On Sun, May 18, 2014 at 03:49:53PM -0400, John D. Verne wrote: On Sun, May 18, 2014 at 03:04:30PM -0400, John D. Verne wrote: I'm staring at the ACPI code trying to figure out the various defines Uh, I actually meant APIC. I need more sleep. Except I didn't. I did meant ACPI. Cripes, I should go take a nap. I suspect this is one of those cases where I can disable some parts of the ACPI via the BIOS or UKC... -- John D. Verne j...@clevermonkey.org
Re: Boot panic on MP amd64 with 5.5, snapshot kernels
On Sun, May 18, 2014 at 03:04:30PM -0400, John D. Verne wrote: I just got a new amd64 box to run OpenBSD on, but it is panicking on boot when I try to run the 5.5 kernel on it. The panic is unknown MPS interrupt trigger 2 somewhere in the acpi code. bsd.rd, bsd and bsd.mp all panic in the same places, as does bsd.rd from the latest amd64 snapshot. I haven't tried i386 yet. I messed around with boot -c and verbose mode, and captured the entire dmseg from ddb from 5.5 /bsd. I've placed the lame camera phone pics here: http://www.clevermonkey.org/OpenBSD/amd64_55_info/ I couldn't figure out how to save the actual text. Sorry. Also, if it matters, this is all done via PXE boot, as this box has no CDROM or floppy, and I didn't have the facility to make an amd64 USB stick (this is my only amd64 box.) Also, if it matters, I opted to not get the Atom version of this board because I wanted some of the more interesting chipset support. It is a Celeron J1000. I also wanted the better supported Intel graphics stuff (even though I suspect this box will run in text-only mode most of its life.) I'm staring at the ACPI code trying to figure out the various defines used, and 2 seems to be there. I welcome conversations about what this code is doing and what it wants out of total curiousity. -- John D. Verne j...@clevermonkey.org Is this 5.5-current? We made some changes in the MADT code recently. -ml
Re: Boot panic on MP amd64 with 5.5, snapshot kernels
On Sun, May 18, 2014 at 01:05:52PM -0700, Mike Larkin wrote: On Sun, May 18, 2014 at 03:04:30PM -0400, John D. Verne wrote: I just got a new amd64 box to run OpenBSD on, but it is panicking on boot when I try to run the 5.5 kernel on it. The panic is unknown MPS interrupt trigger 2 somewhere in the acpi code. bsd.rd, bsd and bsd.mp all panic in the same places, as does bsd.rd from the latest amd64 snapshot. Is this 5.5-current? We made some changes in the MADT code recently. I tried the 18-May-2014 amd64 snapshot. That's about as current as I can get right now. -- John D. Verne j...@clevermonkey.org
Re: Boot panic on MP amd64 with 5.5, snapshot kernels
On Sun, May 18, 2014 at 07:28:04PM -0400, John D. Verne wrote: On Sun, May 18, 2014 at 01:05:52PM -0700, Mike Larkin wrote: On Sun, May 18, 2014 at 03:04:30PM -0400, John D. Verne wrote: I just got a new amd64 box to run OpenBSD on, but it is panicking on boot when I try to run the 5.5 kernel on it. The panic is unknown MPS interrupt trigger 2 somewhere in the acpi code. bsd.rd, bsd and bsd.mp all panic in the same places, as does bsd.rd from the latest amd64 snapshot. Is this 5.5-current? We made some changes in the MADT code recently. I tried the 18-May-2014 amd64 snapshot. That's about as current as I can get right now. I see there is very recent change from about 3hrs ago, which I bet is what you are talking about. I'll wait for the next snapshot I guess. -- John D. Verne j...@clevermonkey.org
Re: Boot panic on MP amd64 with 5.5, snapshot kernels
On Sun, May 18, 2014 at 07:55:44PM -0400, John D. Verne wrote: On Sun, May 18, 2014 at 07:28:04PM -0400, John D. Verne wrote: On Sun, May 18, 2014 at 01:05:52PM -0700, Mike Larkin wrote: On Sun, May 18, 2014 at 03:04:30PM -0400, John D. Verne wrote: I just got a new amd64 box to run OpenBSD on, but it is panicking on boot when I try to run the 5.5 kernel on it. The panic is unknown MPS interrupt trigger 2 somewhere in the acpi code. bsd.rd, bsd and bsd.mp all panic in the same places, as does bsd.rd from the latest amd64 snapshot. Is this 5.5-current? We made some changes in the MADT code recently. I tried the 18-May-2014 amd64 snapshot. That's about as current as I can get right now. I see there is very recent change from about 3hrs ago, which I bet is what you are talking about. I'll wait for the next snapshot I guess. Though, when I look at this change, I don't see how it could help. Given the panic stacktrace we are getting beyond this changed line anyway to acpimadt_cfg_intr() where the panic is thrown. -- John D. Verne j...@clevermonkey.org