Harvey-OS/harvey

Kernel panic during boot in Arch Linux

fhs opened this issue · 3 comments

fhs commented

Expected behavior

Successful bootup

Actual behavior

Kernel panic

Steps to reproduce

  • ./bootstrap.sh
  • CC=gcc ARCH=amd64 util/build
  • export HARVEY=$(pwd)
  • util/GO9PCPU

I'm using Arch Linux, with gcc 8.2.1+20180831, binutils 2.31.1, and qemu 3.0.0.

QEMU output

qemu-system-x86_64 -s -cpu Opteron_G1 -smp 4 -m 2048 -enable-kvm -serial stdio --machine pc,accel=kvm -net nic,model=rtl8139 -net

Harvey
mmuinit: vmstart 0xfffffffff0000000 vmunused 0xfffffffff0b85000 vmunmapped 0xfffffffff0c00000 vmend 0xfffffffff4000000
sys->pd 0x108003 0x108023
cpu0: mmu l3 pte 0xfffffffff0106ff8 = 107023
cpu0: mmu l2 pte 0xfffffffff0107ff8 = 108023
cpu0: mmu l1 pte 0xfffffffff0108c00 = e3
cpu0: mmu l1 pte 0xfffffffff0108c00 = e3
base 0xfffffffff0b8c000 ptr 0xfffffffff0b8c000 nunits 3437569
allocated
 free ok
parsexsdt: tbl fffffffff0b8e04c, end 12
ACPI initialized
	acpiinit();
physalloc color=0 base=0x4000000 size=0x7bfe0000
before mpacpi, maxcores 1024
APIC lapic paddr 0xfee00000, flags 0x00000001
ASlapic 0
apic proc 0/0 apicid 0 
ASlapic 1
apic proc 1/1 apicid 1 
ASlapic 2
apic proc 2/2 apicid 2 
ASlapic 3
apic proc 3/3 apicid 3 
ASioapic 0
ioapic 0 addr fec00000 ibase 0 
Could not find an IOAPIC for global irq 62!
Unable to build IOAPIC route for irq 62
intrenable: couldn't enable irq 62, tbdf 0x5ffff00 for #IPI
apic0: hz 95475120 max 954751 min 9547 div 3
sipiptr 0xfffffffff0003000 sipipa 0x3000
p 0xfffffffff0ba4000 sipiptr[-1] 0xba4000
p 0xfffffffff0bb1000 sipiptr[-1] 0xbb1000
p 0xfffffffff0bbe000 sipiptr[-1] 0xbbe000
ioapicintrinit: success
virtio: initializing
virtio: no devices
bus dev type vid  did intl memory
0   0/0 06 00 00 8086 1237 x   0  
0   1/0 06 01 00 8086 7000 x   0  
0   1/1 01 01 80 8086 7010 x   0  4:0000c101 16 
0   1/3 06 80 00 8086 7113 A   9  
0   2/0 03 00 00 1234 1111 x   0  0:fd000008 16777216 2:febd0000 4096 
0   3/0 02 00 00 10ec 8139 A  11  0:0000c001 256 1:febd1000 256 
ioapicintrinit: success
ioapicintrinit: success
#l0: rtl8139: 100Mbps port 0xc000 irq 11 tu 1514: 525400123456
unknown color for cpu0
cpu1 color -1 role TC tsc 5793392462
cpu3 color -1 role TC tsc 5793392945
schedinit: no sch for cpu1
schedinit: no sch for cpu3
unknown color for cpu1
cpu2 color -1 role TC tsc 5793383055
Scan fffffffff0000000
bootcore: all cores done
Scan fffffffff00f0000
unknown color for cpu3
get_coreboot_info: ret -1
schedinit: no sch for cpu2
virtio-9p initializing
CPU Freq. 266MHz
virtio-serial-pci initializing
unknown color for cpu2
panic: cpu1: boot process died: empty file name
schedinit...
die:wait forever
gmacd commented

I think qemu 3 is the difference here. Since I moved from qemu 2.9 -> 3 I've started seeing the same error.

elbing commented

Same issue than in debian testing distro.

gmacd commented

Fixed with latest code. Run ./util/build clean.json before rebuilding.

  • 👍1
  • 🎉1