6.0-BETA2 STRANGE PROBLEMS

[Giovanni P. Tirloni]

guilherme wrote:
> Hi, I think that developers would polish 6.0 more before release it so 
> fast.
> Here's the output of dmesg. I have got serious problems testing this BETA2.

  FreeBSD 6.0 wasn't released yet. This is a BETA version used for 
testing. We will have some more betas and then the release candidates.

> 
> - It crashed in install, outing to db> prompt.
> - The boot take a year, because of ACPI problems.
> - Strange VM errors.

  [snip]

> lock order reversal
>  1st 0xc097b240 UMA lock (UMA lock) @ /usr/src/sys/vm/uma_core.c:1494
>  2nd 0xc1060144 system map (system map) @ /usr/src/sys/vm/vm_map.c:2317
> KDB: stack backtrace:
> kdb_backtrace(0,ffffffff,c0930358,c0930498,c08bace4) at kdb_backtrace+0x29
> witness_checkorder(c1060144,9,c08714b6,90d) at witness_checkorder+0x564
> _mtx_lock_flags(c1060144,0,c08714b6,90d) at _mtx_lock_flags+0x5b
> _vm_map_lock(c10600c0,c08714b6,90d) at _vm_map_lock+0x26
> vm_map_remove(c10600c0,c1cc6000,c1cc7000,c8398c08,c077ee01) at 
> vm_map_remove+0x1f
> kmem_free(c10600c0,c1cc6000,1000,c8398c38,c077e7ae) at kmem_free+0x25
> page_free(c1cc6000,1000,2) at page_free+0x29
> zone_drain(c104a960) at zone_drain+0x26a
> zone_foreach(c077e544,c8398cec,c079062f,c1341c00,c8398c74) at 
> zone_foreach+0x37
> uma_reclaim(c1341c00,c8398c74,0,c0926980,c8398c80) at uma_reclaim+0x12
> vm_pageout_scan(0,c097b6a0,0,c08729a3,5c3) at vm_pageout_scan+0x103
> vm_pageout(0,c8398d38,0,c07913e8,0) at vm_pageout+0x2c3
> fork_exit(c07913e8,0,c8398d38) at fork_exit+0xa0
> fork_trampoline() at fork_trampoline+0x8
> --- trap 0x1, eip = 0, esp = 0xc8398d6c, ebp = 0 ---

You are running a kernel full of debugging options and it has detected a 
lock order reversal and sent you to the debugger so you can get more 
information about what's happening.

A LOR [1] isn't a serious problem, AFAIK, so you can just type 
"continue" and the system will hopefully continue booting.

BTW, if after installing you feel the system is a bit slow compared to 
what you're used to read the beginning of /usr/src/UPDATING. You have to 
remove the debug options and recompile the kernel because they will 
cause speed to drop.

As for ACPI, there have been some problem but it seems to have 
stabilized for a while. You can' try to disable it.

[1]- http://kerneltrap.org/node/3620
[2]- http://sources.zabbadoz.net/freebsd/lor.html

Good luck,

-- 
Giovanni P. Tirloni / gpt at tirloni.org