Subject: Re: serial cable / Intel SR1500AL
To: None <port-i386@netbsd.org>
From: Patrick Welche <prlw1@newn.cam.ac.uk>
List: port-i386
Date: 07/05/2007 18:51:05
On Tue, Jul 03, 2007 at 03:19:40PM +0100, Patrick Welche wrote:
> An Intel SR1500AL box with 4GB ram has just arrived. Booting GENERIC
> gets me a panic in pci_map_reg. If I disable acpi, I get the attached
> dmesg, which is odd as less ram is detected. Sadly it seems that the
> raid controler isn't supported:
> 
> Intel product 0x2682 (RAID mass storage, revision 0x09) at pci0 dev 31 function 
> 2 not configured
> 
> I thought I would try to make some in-roads, but am stuck at the first
> hurdle: trying to make use of the serial port. I made a cable with pins:
> 
> RJ45            9-pin male
> 1 DSR, 2 DCD -> 4 DTR
> 3 DTR        -> 1 DCD, 6 DSR
> 4 GND        -> 5 GND
> 5 RxD        -> 3 TxD
> 6 TxD        -> 2 RxD
> 7 CTS        -> 7 RTS
> 8 RTS        -> 8 CTS
> 
> Does that make sense?

I resoldered the cable now trying the "Yost Serial Device Wiring Standard"
instead of EIA/TIA 561:

RJ45     9-pin male
1 RTS -> 8 CTS
2 DTR -> 6 DSR
3 TxD -> 2 RxD
4 GND -> 5 GND
5 GND -> 5 GND (yes 4 and 5 to 5)
6 RxD -> 3 TxD
7 DSR -> 4 DTR
8 CTS -> 7 RTS

and it works! I now have serial console, and it is com1, 0x2f8.


The box returned from its makers after "CPU thermal event" trouble with a

amr0 at pci7 dev 14 function 0: AMI RAID <LSI MegaRAID SATA300-8X PCI-X>
amr0: interrupting at irq 11
amr0: firmware 815C, BIOS H432, 128MB RAM
ld0 at amr0 unit 0: RAID 1, optimal
ld0: 465 GB, 60788 cyl, 255 head, 63 sec, 512 bytes/sect x 976562176 sectors

which on the one hand is good for getting the box running, but on the other
not so good for trying to get the Intel RAID working.


Not sure what PAE support means, but it seems NetBSD/i386 doesn't have it
(there is 4GB of memory):

>> NetBSD/i386 PXE Boot, Revision 1.1
>> Memory: 557/4187136 k
6643044+5379012+289272 [332624+302569]=0xc5a24c
WARNING: skipping large memory map entry: 0x100000000/0x60000000/0x1
total memory = 2554 MB
avail memory = 2494 MB

which seems to be related to the following in the BIOS:

Memory Mapped I/O Start Address 2.5GB
Select the start of the reserved memory region for PCI memory mapped I/O space
that ends at 4GB.
Warning: Depending on the system configuration, this option may impact the
amount of system memory detected by an OS without Physical Address Extension
(PAE) support.



If acpi is not disabled, I get a panic:

RTC BIOS diagnostic error a<fixed_disk>
SMBIOS rev. 2.5 @ 0x9fa42000 (70 entries)
mainbus0 (root)
cpu0 at mainbus0 apid 0: (boot processor)
cpu0: Intel Core 2 (Merom) (686-class), 1995.15 MHz, id 0x6f6
cpu0: "Intel(R) Xeon(R) CPU            5130  @ 2.00GHz"
cpu1 at mainbus0 apid 6: (application processor)
cpu1: not started
cpu2 at mainbus0 apid 1: (application processor)
cpu2: not started
cpu3 at mainbus0 apid 7: (application processor)
cpu3: not started
ioapic0 at mainbus0 apid 8
ioapic1 at mainbus0 apid 9
acpi0 at mainbus0: Advanced Configuration and Power Interface
ACPI-Fast 24-bit timer
panic: pci_make_tag: bad request

cpu_Debugger(c427d1e0,c050a3bf,0,c0b167e4,1) at netbsd:cpu_Debugger+0x4
panic(c050a6f9,0,c0b16808,1,c0b16824) at netbsd:panic+0x12b
pci_make_tag(c0b16824,0,b,ff,0) at netbsd:pci_make_tag+0x4a
mpacpi_pcibus_cb(c427d1e0,3,c4279800,0,1ab5f00) at netbsd:mpacpi_pcibus_cb+0x29d
AcpiNsWalkNamespace(6,c3ab5de0,64,1,c02e28da) at netbsd:AcpiNsWalkNamespace+0xf4
AcpiWalkNamespace(6,c3ab5de0,64,c02e28da,c4279800) at netbsd:AcpiWalkNamespace+0x46
mpacpi_find_interrupts(c4279800,a,a,c4279858,92e617) at netbsd:mpacpi_find_interrupts+0xc5
acpi_md_callback(c4279800,c4279800,c0b16ac8,c03c80d0,5) at netbsd:acpi_md_callback+0xf
acpi_attach(c427bfc0,c4279800,c0b16b4c,0,1) at netbsd:acpi_attach+0x25e
config_attach_loc(c427bfc0,c0547dd0,0,c0b16b4c,0) at netbsd:config_attach_loc+0x12d
config_found_ia(c427bfc0,c04e42f6,c0b16b4c,0,e) at netbsd:config_found_ia+0x1b
mainbus_attach(0,c427bfc0,0,61,0) at netbsd:mainbus_attach+0x103
config_attach_loc(0,c0547338,0,0,0) at netbsd:config_attach_loc+0x12d
config_attach(0,c0547338,0,0,c0b13010) at netbsd:config_attach+0x19
config_rootfound(c04e3d5f,0,c0b7aff0,c0b7aff0,b13000) at netbsd:config_rootfound+0x27
cpu_configure(c0a4eb40,c0b13010,c0b16c78,c02809d7,0) at netbsd:cpu_configure+0x1e
configure(0,0,0,0,0) at netbsd:configure+0x4a
main(fbff,c01002ac,0,0,0) at netbsd:main+0x109


And if acpi is disabled, booting seems to simply hang after

cd0 at atapibus0 drive 1: <DV-28E-R, , 1.8A> cdrom removable

When trying kgdb, I get
(gdb) target remote /dev/tty00
Remote debugging using /dev/tty00
0xc03a247d in dvd_read_physical (cd=0xc05e5050, s=0x0)
    at ../../../../dev/scsipi/cd.c:2222
2222            error = scsipi_command(cd->sc_periph, &cmd, 12, bf, sizeof(bf),
(gdb) c
Continuing.

Program received signal SIGTRAP, Trace/breakpoint trap.
0xc03a247d in dvd_read_physical (cd=0xc0d5e784, s=0xe)
    at ../../../../dev/scsipi/cd.c:2222
2222            error = scsipi_command(cd->sc_periph, &cmd, 12, bf, sizeof(bf),
(gdb) c
Continuing.
{hangs]


Anyone interested in any of this? Is it worth splitting into PRs?

Cheers,

Patrick