kernel/1023: kernel panic in ppp code

[Urazmetov Wasilx <wasilx@forest.ihep.su>]

>Number:         1023
>Category:       kernel
>Synopsis:       kernel panic in ppp code
>Confidential:   no
>Severity:       serious
>Priority:       medium
>Responsible:    bugs
>State:          open
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Fri Dec 24 03:50:01 MST 1999
>Last-Modified:
>Originator:     Urazmetov Wasilx
>Organization:
IHEP
>Release:        -current
>Environment:
	System      : OpenBSD 2.6
	Architecture: OpenBSD.i386
	Machine     : i486
>Description:
	OpenBSD kernel crashes under medium (about 2) load in kernel ppp code:
|
| ppp0: bad fcs 4457
| z_decompress0: bad seq # 1689, expected 1683
| ppp0: decompress failed 1
| vm_fault (0xe0212b1c, e0a04000, 1, 0) -> 1
| kernel: page fault trap, code=0
| Stopped at      _deflate_slow+0x2b:     movzbl  0x2(%eax,%ecx,1),%eax
| ddb> 
|
   Note.1 I have `ddb.panic=0' in /etc/sysctl.conf and
          `sysctl ddb.panic' does confirm the setting been applied.
          Why does the kernel stop in ddb then?

   Note.2 OpenBSD kernel PPP code appears to be buggy as it claims to get
          lots of corrupted packets from serial line.
          I have tested the hardware used (the computer and internal modems)
          under Linux, it is ok.

   Note.3 PPP is configured to use BSD compression.
          I've tried `deflate' with the same success:
          lots of ``corrupted'' data.

   Note.4 These panics mostly happen under some load
          (probably the main matter is intensive paging),
          e.g. system build running increases crash probability dramatically.

--- dmesg begin ---------------------------------------------------------------
bootargv: diskinfo 0xe033d00c cksumlen 1 memmap 0xe033d088 apminfo 0xe033d0d0
OpenBSD 2.6-current (kernel) #22: Mon Dec 20 20:05:20 MSK 1999
    root@dclxvi:/home/dclxvi/kernel
cpu0: Intel 486DX4 ("GenuineIntel" 486-class)
cpu0: FPU,V86
BIOS mem  = 655360 conventional, 15728640 extended
real mem  = 16384000
avail mem = 13168640
using 225 buffers containing 921600 bytes of memory
mainbus0 (root)
bios0 at mainbus0: AT/286+(00) BIOS, date 07/25/94
apm at bios0 function 0x15 not configured
isa0 at mainbus0
isadma0 at isa0
wdc0 at isa0 port 0x1f0/8 irq 14
wd0 at wdc0 channel 0 drive 0: <Conner Peripherals 1275MB - CFS1275A>
wd0: can use 16-bit, PIO mode 4
wd0: 16-sector PIO, LBA, 1219MB, 2477 cyl, 16 head, 63 sec, 2496876 sectors
wd1 at wdc0 channel 0 drive 1: <Conner Peripherals 1275MB - CFA1275A>
wd1: can use 16-bit, PIO mode 4
wd1: 128-sector PIO, LBA, 1220MB, 2479 cyl, 16 head, 63 sec, 2499030 sectors
ne0 at isa0 port 0x300/32 irq 10
ne0: NE2000 Ethernet
ne0: address 00:00:01:29:58:68
ne1 at isa0 port 0x320/32 irq 11
ne1: NE2000 Ethernet
ne1: address 00:20:78:07:94:6f
pcppi0 at isa0 port 0x61
sysbeep0 at pcppi0
npx0 at isa0 port 0xf0/16: using exception 16
pccom0 at isa0 port 0x3f8/8 irq 3: ns16450, no fifo
pccom0: console
pccom1 at isa0 port 0x2f8/8 irq 4: ns16450, no fifo
pccom2 at isa0 port 0x3e8/8 irq 5: ns16550a, 16 byte fifo
pccom3 at isa0 port 0x2e8/8 irq 7: ns16550a, 16 byte fifo
vt0 at isa0 port 0x60/16 irq 1: vga 80/132 col, color, 8 scr, mf2-kbd
biomask 4000 netmask 4c00 ttymask 4c02
pctr: no performance counters in CPU
dkcsum: wd0 matched BIOS disk 80
dkcsum: wd1 matched BIOS disk 81
root on wd0a
rootdev=0x0 rrootdev=0x300 rawdev=0x302
WARNING: / was not properly unmounted
IP Filter: initialized.  Default = pass all, Logging = enabled
IP Filter: v3.3.5
--- dmesg   end ---------------------------------------------------------------

>How-To-Repeat:
	Get ix86 host, modems, install obsd, configure pppd,
	run `make build' in /usr/src/ (running sshd and squid is desirable),
	launch PPP connection to the host via serial line uisng pppd...
>Fix:
	

>Audit-Trail:
>Unformatted: