Bug 72128 - 5.3-beta6 doesn't boot
Summary: 5.3-beta6 doesn't boot
Status: Closed FIXED
Alias: None
Product: Base System
Classification: Unclassified
Component: ia64 (show other bugs)
Version: Unspecified
Hardware: Any Any
: Normal Affects Only Me
Assignee: freebsd-ia64 (Nobody)
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2004-09-27 14:20 UTC by Jens Axboe
Modified: 2005-07-03 02:10 UTC (History)
0 users

See Also:


Attachments
bsd53.txt (31.13 KB, text/plain)
2004-10-18 20:36 UTC, Arun Sharma
no flags Details
bsd53.txt (31.13 KB, text/plain)
2004-10-18 20:36 UTC, Arun Sharma
no flags Details
file.dat (212 bytes, text/plain; charset="us-ascii")
2004-10-18 20:36 UTC, Arun Sharma
no flags Details

Note You need to log in before you can comment on or make changes to this bug.
Description Jens Axboe 2004-09-27 14:20:24 UTC
I've tried beta3/4/5/6 and they all hang right after printing:

Entering /boot/kernel/kernel at 0xe000000004058000...

How-To-Repeat: Just boot the iso.
Comment 1 Jens Axboe 2004-10-04 09:28:13 UTC
Hi,

Is there any progress here? Do you need more info?

-- 
Jens Axboe
Comment 2 Mark Linimon freebsd_committer freebsd_triage 2004-10-08 06:54:38 UTC
Responsible Changed
From-To: freebsd-bugs->freebsd-ia64

This may be processor-specific.
Comment 3 Arun Sharma 2004-10-08 22:24:00 UTC
On 10/7/2004 11:02 PM, Mark Linimon wrote:

> This may be processor-specific.
> 
> http://www.freebsd.org/cgi/query-pr.cgi?pr=72128

I just tested that the 5.3-beta7 works fine on a 4 way Tiger with 1GB RAM. I don't think there have been critical bug fixes between beta6 and beta7 that affect the tiger. Also, on a 4GB machine, FreeBSD ignores RAM above 1GB.

Jens, were you using a serial console ? FreeBSD doesn't support a VGA console yet on ia64.

	-Arun
Comment 4 Arun Sharma 2004-10-09 01:44:56 UTC
On 10/8/2004 2:24 PM, Arun Sharma wrote:
> On 10/7/2004 11:02 PM, Mark Linimon wrote:
> 
>> This may be processor-specific.
>>
>> http://www.freebsd.org/cgi/query-pr.cgi?pr=72128
> 
> 
> I just tested that the 5.3-beta7 works fine on a 4 way Tiger with 1GB 
> RAM. I don't think there have been critical bug fixes between beta6 and 
> beta7 that affect the tiger. Also, on a 4GB machine, FreeBSD ignores RAM 
> above 1GB.

I meant to say FreeBSD ignores RAM above 4GB physical. So depending on your memory map, this may or may not be a problem.

	-Arun
Comment 5 Jens Axboe 2004-10-09 19:01:33 UTC
On Fri, Oct 08 2004, Arun Sharma wrote:
> On 10/7/2004 11:02 PM, Mark Linimon wrote:
> 
> >This may be processor-specific.
> >
> >http://www.freebsd.org/cgi/query-pr.cgi?pr=72128
> 
> I just tested that the 5.3-beta7 works fine on a 4 way Tiger with 1GB RAM. 
> I don't think there have been critical bug fixes between beta6 and beta7 
> that affect the tiger. Also, on a 4GB machine, FreeBSD ignores RAM above 
> 1GB.
> 
> Jens, were you using a serial console ? FreeBSD doesn't support a VGA 
> console yet on ia64.

I am. I did attach a monitor as well for testing since I didn't know if
freebsd utilized the serial console automatically. I since looked at the
boot loader config and saw that it did and the config was correct.

-- 
Jens Axboe
Comment 6 Jens Axboe 2004-10-09 19:02:26 UTC
On Fri, Oct 08 2004, Arun Sharma wrote:
> On 10/8/2004 2:24 PM, Arun Sharma wrote:
> >On 10/7/2004 11:02 PM, Mark Linimon wrote:
> >
> >>This may be processor-specific.
> >>
> >>http://www.freebsd.org/cgi/query-pr.cgi?pr=72128
> >
> >
> >I just tested that the 5.3-beta7 works fine on a 4 way Tiger with 1GB 
> >RAM. I don't think there have been critical bug fixes between beta6 and 
> >beta7 that affect the tiger. Also, on a 4GB machine, FreeBSD ignores RAM 
> >above 1GB.
> 
> I meant to say FreeBSD ignores RAM above 4GB physical. So depending on your 
> memory map, this may or may not be a problem.

I can verify my memory map in Linux, I'll send it in on the next boot.

-- 
Jens Axboe
Comment 7 Jens Axboe 2004-10-18 12:34:30 UTC
On Sat, Oct 09 2004, Jens Axboe wrote:
> On Fri, Oct 08 2004, Arun Sharma wrote:
> > On 10/8/2004 2:24 PM, Arun Sharma wrote:
> > >On 10/7/2004 11:02 PM, Mark Linimon wrote:
> > >
> > >>This may be processor-specific.
> > >>
> > >>http://www.freebsd.org/cgi/query-pr.cgi?pr=72128
> > >
> > >
> > >I just tested that the 5.3-beta7 works fine on a 4 way Tiger with 1GB 
> > >RAM. I don't think there have been critical bug fixes between beta6 and 
> > >beta7 that affect the tiger. Also, on a 4GB machine, FreeBSD ignores RAM 
> > >above 1GB.
> > 
> > I meant to say FreeBSD ignores RAM above 4GB physical. So depending on your 
> > memory map, this may or may not be a problem.
> 
> I can verify my memory map in Linux, I'll send it in on the next boot.

Sorry about the delay, was away on business.

These are the only comments that Linux gives about the memory layout:

efi.trim_top: ignoring 4KB of memory at 0x0 due to granule hole at 0x0
efi.trim_top: ignoring 24KB of memory at 0x1000 due to granule hole at 0x0
efi.trim_top: ignoring 8KB of memory at 0x7000 due to granule hole at 0x0
efi.trim_top: ignoring 484KB of memory at 0x9000 due to granule hole at 0x0
efi.trim_top: ignoring 4KB of memory at 0x84000 due to granule hole at 0x0
efi.trim_top: ignoring 108KB of memory at 0x85000 due to granule hole at 0x0
efi.trim_bottom: ignoring 15360KB of memory at 0x100000 due to granule hole at 0x0
efi.trim_bottom: ignoring 7168KB of memory at 0x7d900000 due to granule hole at 0x7d000000

Virtual mem_map starts at 0xa0007fffffc70000
On node 0 totalpages: 64916
  DMA zone: 64916 pages, LIFO batch:4
  Normal zone: 0 pages, LIFO batch:1
  HighMem zone: 0 pages, LIFO batch:1
Built 1 zonelists

Let me know if you need more info!

-- 
Jens Axboe
Comment 8 Arun Sharma 2004-10-18 20:36:56 UTC
On 10/18/2004 4:34 AM, Jens Axboe wrote:
> On Sat, Oct 09 2004, Jens Axboe wrote:
>> On Fri, Oct 08 2004, Arun Sharma wrote:
>> > On 10/8/2004 2:24 PM, Arun Sharma wrote:
>> > >On 10/7/2004 11:02 PM, Mark Linimon wrote:
>> > >
>> > >>This may be processor-specific.
>> > >>
>> > >>http://www.freebsd.org/cgi/query-pr.cgi?pr=72128
>> > >
>> > >
>> > >I just tested that the 5.3-beta7 works fine on a 4 way Tiger with 1GB 
>> > >RAM. I don't think there have been critical bug fixes between beta6 and 
>> > >beta7 that affect the tiger. Also, on a 4GB machine, FreeBSD ignores RAM 
>> > >above 1GB.
>> > 
>> > I meant to say FreeBSD ignores RAM above 4GB physical. So depending on your 
>> > memory map, this may or may not be a problem.
>> 
>> I can verify my memory map in Linux, I'll send it in on the next boot.
> 
> Sorry about the delay, was away on business.
> 
> These are the only comments that Linux gives about the memory layout:
> 

I can't reproduce your problem on a tiger4 with 4GB RAM. Attached is the memory map obtained via memmap EFI command and the dmesg.

	-Arun
Comment 9 Arun Sharma 2004-10-18 20:36:56 UTC
On 10/18/2004 4:34 AM, Jens Axboe wrote:
> On Sat, Oct 09 2004, Jens Axboe wrote:
>> On Fri, Oct 08 2004, Arun Sharma wrote:
>> > On 10/8/2004 2:24 PM, Arun Sharma wrote:
>> > >On 10/7/2004 11:02 PM, Mark Linimon wrote:
>> > >
>> > >>This may be processor-specific.
>> > >>
>> > >>http://www.freebsd.org/cgi/query-pr.cgi?pr=72128
>> > >
>> > >
>> > >I just tested that the 5.3-beta7 works fine on a 4 way Tiger with 1GB 
>> > >RAM. I don't think there have been critical bug fixes between beta6 and 
>> > >beta7 that affect the tiger. Also, on a 4GB machine, FreeBSD ignores RAM 
>> > >above 1GB.
>> > 
>> > I meant to say FreeBSD ignores RAM above 4GB physical. So depending on your 
>> > memory map, this may or may not be a problem.
>> 
>> I can verify my memory map in Linux, I'll send it in on the next boot.
> 
> Sorry about the delay, was away on business.
> 
> These are the only comments that Linux gives about the memory layout:
> 

I can't reproduce your problem on a tiger4 with 4GB RAM. Attached is the memory map obtained via memmap EFI command and the dmesg.

	-Arun
Comment 10 Jens Axboe 2004-10-20 09:29:52 UTC
Hi,

Here's the memmap on my box:

Shell> memmap

Type       Start            End               # Pages
Attributes
BS_data    0000000000000000-0000000000000FFF  0000000000000001
0000000000000009
available  0000000000001000-0000000000006FFF  0000000000000006
0000000000000009
BS_data    0000000000007000-0000000000008FFF  0000000000000002
0000000000000009
available  0000000000009000-0000000000081FFF  0000000000000079
0000000000000009
RT_data    0000000000082000-0000000000083FFF  0000000000000002
8000000000000009
available  0000000000084000-0000000000084FFF  0000000000000001
0000000000000009
BS_data    0000000000085000-000000000009FFFF  000000000000001B
0000000000000009
RT_code    00000000000C0000-00000000000FFFFF  0000000000000040
8000000000000009
available  0000000000100000-000000000FF7FFFF  000000000000FE80
000000000000000B
BS_data    000000000FF80000-000000000FFFFFFF  0000000000000080
000000000000000B
available  0000000010000000-000000007D8FFFFF  000000000006D900
000000000000000B
BS_code    000000007D900000-000000007F97FFFF  0000000000002080
000000000000000B
available  000000007F980000-000000007F9FFFFF  0000000000000080
000000000000000B
RT_code    000000007FA00000-000000007FDFFFFF  0000000000000400
8000000000000009
PAL_code   000000007FE00000-000000007FE3FFFF  0000000000000040
8000000000000009
RT_code    000000007FE40000-000000007FE97FFF  0000000000000058
8000000000000009
available  000000007FE98000-000000007FEE5FFF  000000000000004E
000000000000000B
BS_data    000000007FEE6000-000000007FF08FFF  0000000000000023
000000000000000B
available  000000007FF09000-000000007FF27FFF  000000000000001F
000000000000000B
BS_data    000000007FF28000-000000007FF2FFFF  0000000000000008
000000000000000B
Shell>     000000007FF30000-000000007FFFFFFF  00000000000000D0
8000000000000009
MemMapIO   00000000FE000000-00000000FEFFFFFF  0000000000001000
0000000000000001
RT_data    00000000FF000000-00000000FFFFFFFF  0000000000001000
8000000000000001
available  0000000180000000-00000001FEFAFFFF  000000000007EFB0
000000000000000B
LoaderData 00000001FEFB0000-00000001FEFB7FFF  0000000000000008
000000000000000B
LoaderCode 00000001FEFB8000-00000001FEFFDFFF  0000000000000046
000000000000000B
available  00000001FEFFE000-00000001FF453FFF  0000000000000456
000000000000000B
BS_data    00000001FF454000-00000001FF800FFF  00000000000003AD
000000000000000B
available  00000001FF801000-00000001FF88BFFF  000000000000008B
000000000000000B
BS_data    00000001FF88C000-00000001FF9F9FFF  000000000000016E
000000000000000B
available  00000001FF9FA000-00000001FF9FAFFF  0000000000000001
000000000000000B
BS_data    00000001FF9FB000-00000001FF9FDFFF  0000000000000003
000000000000000B
available  00000001FF9FE000-00000001FFD7FFFF  0000000000000382
000000000000000B
BS_code    00000001FFD80000-00000001FFDFDFFF  000000000000007E
000000000000000B
available  00000001FFDFE000-00000001FFE11FFF  0000000000000014
000000000000000B
RT_code    00000001FFE12000-00000001FFE7DFFF  000000000000006C
8000000000000009
available  00000001FFE7E000-00000001FFFB5FFF  0000000000000138
000000000000000B
RT_data    00000001FFFB6000-00000001FFFFFFFF  000000000000004A
8000000000000009
MemMapIO   00000FFFF8000000-00000FFFFBFFFFFF  0000000000004000
8000000000000001
MemPortIO  00000FFFFC000000-00000FFFFFFFFFFF  0000000000004000
8000000000000001

  LoaderCode:      70 Pages (286,720)
  LoaderData:       8 Pages (32,768)
  BS_code   :   8,446 Pages (34,594,816)
  BS_data   :   1,511 Pages (6,189,056)
  RT_code   :   1,284 Pages (5,259,264)
  RT_data   :   4,380 Pages (17,940,480)
  available : 1,036,877 Pages (4,247,048,192)
  MemMapIO  :  20,480 Pages (83,886,080)
  MemPortIO :  16,384 Pages (67,108,864)
  PAL_code  :      64 Pages (262,144)
Total Memory: 4,111 MB (4,311,613,440) Bytes

-- 
Jens Axboe
Comment 11 Arun Sharma 2004-10-21 19:45:40 UTC
Jens Axboe wrote:
> Hi,
> 
> Here's the memmap on my box:
> 
> Shell> memmap

Looks pretty similar to my 4GB box. It looks like your serial console is 
working fine for EFI. Could you please check the settings for the kernel 
in the boot loader ?

I use something like:

hw.uart.console="io:0x2f8,br:57600"

Your baud rate might be different. But you certainly need to change the 
io to 0x2f8 (COM2) from the default of 0x3f8 (COM1) on the Tiger.

	-Arun
Comment 12 Jens Axboe 2004-10-21 21:21:06 UTC
On Thu, Oct 21 2004, Arun Sharma wrote:
> Jens Axboe wrote:
> >Hi,
> >
> >Here's the memmap on my box:
> >
> >Shell> memmap
> 
> Looks pretty similar to my 4GB box. It looks like your serial console is 
> working fine for EFI. Could you please check the settings for the kernel 
> in the boot loader ?
> 
> I use something like:
> 
> hw.uart.console="io:0x2f8,br:57600"
> 
> Your baud rate might be different. But you certainly need to change the 
> io to 0x2f8 (COM2) from the default of 0x3f8 (COM1) on the Tiger.

That might just be it, I actually suspected this earlier today. But I
thought the serial port on the tiger _was_ on 0x3f8. I'll give it a shot
tomorrow, thanks.

-- 
Jens Axboe
Comment 13 Marcel Moolenaar 2004-10-31 04:53:15 UTC
Was it the console setting and if yes, can this PR be closed?
Thanks,

-- 
  Marcel Moolenaar         USPA: A-39004          marcel@xcllnt.net
Comment 14 Marcel Moolenaar freebsd_committer freebsd_triage 2004-11-02 06:29:18 UTC
State Changed
From-To: open->closed

The PR describes the typical case of not using a serial console or 
not having the serial console setup correctly. Given that the box 
does not create the DIG64 HCDP table, there's no other way to tell 
the kernel which serial port to use than setting the hw.uart.console 
variable.