panic at serial boot

[mh at kernel32.de]

Am 2013-03-27 14:32, schrieb Konstantin Belousov:
> On Wed, Mar 27, 2013 at 01:56:29PM +0100, mh at kernel32.de wrote:
>> Am 2013-03-27 06:54, schrieb Andriy Gapon:
>> > on 27/03/2013 00:00 mh at kernel32.de said the following:
>> >> Hi Ho,
>> >>
>> >> I'm trying to install FreeBSD 10.0-CURRENT snapshot from 
>> 23.03.2013
>> >> on a Dell
>> >> M620 blade.
>> >> However, I'm getting a panic and drop to db while booting.
>> >>
>> >> Looks like that:
>> >>
>> >> Root mount waiting for: usbus1 usbus0
>> >> ugen0.2: <vendor 0x8087> at usbus0
>> >> uhub2: <vendor 0x8087 product 0x0024, class 9/0, rev 2.00/0.00, 
>> addr
>> >> 2> on usbus0
>> >> ugen1.2: <vendor 0x8087> at usbus1
>> >> uhub3: <vendor 0x8087 product 0x0024, class 9/0, rev 2.00/0.00, 
>> addr
>> >> 2> on usbus1
>> >> Root mount waiting for: usbus1 usbus0
>> >> uhub2: 6 ports with 6 removable, self powered
>> >> uhub3: 8 ports with 8 removable, self powered
>> >> ugen0.3: <no manufacturer> at usbus0
>> >> uhub4: <no manufacturer Gadget USB HUB, class 9/0, rev 2.00/0.00,
>> >> addr 3> on usbus0
>> >> Root mount waiting for: usbus1 usbus0
>> >> ugen1.3: <Avocent> at usbus1
>> >> ukbd0: <EP1 Interrupt> on usbus1
>> >> kbd0 at ukbd0
>> >> kbd0: ukbd0, generic (0), config:0x0, flags:0x3d0000
>> >> uhub4: 6 ports with 6 removable, self powered
>> >> Root mount waiting for: usbus0
>> >> ugen0.4: <Avocent> at usbus0
>> >> ukbd1: <Keyboard> on usbus0
>> >> kbd2 at ukbd1
>> >> kbd2: ukbd1, generic (0), config:0x0, flags:0x3d0000
>> >> Trying to mount root from ufs:/dev/md0 []...
>> >> start_init: trying /sbin/init
>> >> panic: vm_fault: fault on nofault entry, addr: ffffff80002be000
>> >> cpuid = 15
>> >> KDB: enter: panic
>> >> [ thread pid 4 tid 100101 ]
>> >> Stopped at      kdb_enter+0x3e: movq    $0,kdb_why
>> >> db>
>> >>
>> >> Any idea where to go from here?
>> >>
>> >> initially I tried 9.1-STABLE snapshot from the same date, but 
>> that
>> >> one doesn't
>> >> even give serial output while booting. odd thing, but that's 
>> another
>> >> story.
>> >>
>> >> I just thought since I'm giving -current on this hardware a short
>> >> and there is a
>> >> problem, it might be worth reporting.
>> >> Anything I should do?
>> >
>> > 'bt' command at the 'db>' prompt
>>
>> Here we go:
>>
>> db> bt
>> Tracing pid 5 tid 100109 td 0xfffffe00466bf920
>> kdb_enter() at kdb_enter+0x3e/frame 0xffffff8688d3e340
>> vpanic() at vpanic+0x147/frame 0xffffff8688d3e380
>> panic() at panic+0x43/frame 0xffffff8688d3e3e0
>> vm_fault_hold() at vm_fault_hold+0x14ac/frame 0xffffff8688d3e630
>> vm_fault() at vm_fault+0x77/frame 0xffffff8688d3e670
>> trap_pfault() at trap_pfault+0x207/frame 0xffffff8688d3e720
>> trap() at trap+0x659/frame 0xffffff8688d3e930
>> calltrap() at calltrap+0x8/frame 0xffffff8688d3e930
>> --- trap 0xc, rip = 0xffffffff80c1a396, rsp = 0xffffff8688d3e9f0, 
>> rbp =
>> 0xffffff8688d3ea20 ---
>> bcopy() at bcopy+0x16/frame 0xffffff8688d3ea20
>> md_kthread() at md_kthread+0x17d/frame 0xffffff8688d3ea70
>> fork_exit() at fork_exit+0x84/frame 0xffffff8688d3eab0
>> fork_trampoline() at fork_trampoline+0xe/frame 0xffffff8688d3eab0
>> --- trap 0, rip = 0, rsp = 0xffffff8688d3eb70, rbp = 0 ---
>> db>
>>
>> where to go from now? :)
>
> Do you use preload md(4) ? If yes, try the following patch:
>

Yip. I do. Well, mfsbsd does. I just downloaded the snapshot and gave 
it to mfsbsd to build me an iso image. Transformed that into a pxe 
bootable hd image.
I'm going to checkout -CURRENT sources, apply your patch and let mfsbsd 
build a new image.

Not sure whether time permits that today, but I'm starting right away.

Thanks so far. I will report back :)

./Marian

> diff --git a/sys/dev/md/md.c b/sys/dev/md/md.c
> index d32ddb3..e871d8f 100644
> --- a/sys/dev/md/md.c
> +++ b/sys/dev/md/md.c
> @@ -1008,7 +1008,15 @@ mdinit(struct md_s *sc)
>  	pp = g_new_providerf(gp, "md%d", sc->unit);
>  	pp->mediasize = sc->mediasize;
>  	pp->sectorsize = sc->sectorsize;
> -	pp->flags |= G_PF_ACCEPT_UNMAPPED;
> +	switch (sc->type) {
> +	case MD_MALLOC:
> +	case MD_VNODE:
> +	case MD_SWAP:
> +		pp->flags |= G_PF_ACCEPT_UNMAPPED;
> +		break;
> +	case MD_PRELOAD:
> +		break;
> +	}
>  	sc->gp = gp;
>  	sc->pp = pp;
>  	g_error_provider(pp, 0);