Re: SiS 962/963 support?

Richard Morris wrote:
> Søren Schmidt wrote:
> 
>> Rick Morris wrote:
>>
>>> None of the betas for 5.3 will boot on my systems with the SiS 962 IDE
>>> controllers. I notice that they are not listed in the Handbook under 
>>> supported
>>> hardware. Can anyone tell me whether there are plans to support this
>>> chipset, and if so, what the status is? I've found various posts in 
>>> -bugs and other lists mentioning this chipset in the past, but there 
>>> seems to be no discussion of this for 5.x. All of our deployed units 
>>> with 5.2.1 boot fine with this chipset, but 5.3 freezes on the ado: 
>>> and acdo: messages in boot. Even CTRL-ALT-DEL doesn't work, requiring 
>>> a manual shutdown.
>>>
>>> If anyone has worked on this problem, my company would be very 
>>> interested in talking with you.
>>
>>
>>
>> The SiS southbridges should be supported, at least the two systems I 
>> have here in the lab works just dandy.
>> Could you mail me the output of a verbose boot from 5.2.1 and as far 
>> as it gets on 5.3beta7 please ?
> 
> 
> I have attached the verbose output from 5.2.1, and here is how the 
> output from 5.3beta7 ends (curing CD boot from 
> 5.3-BETA7-i386-bootonly.iso):
> 
> ata0-master: pio=0x0c wdma=0x22 udma=0x45 cable=80pin
> ata0-master: setting PIO4 on SiS 5513 chip
> ata0-master: setting UDMA100 on SiS 5513 chip
> GEOM: create disk ad0 dp=0xc2cb4560
> ad0: <WDC WD400BB-22HEA0/13.03G13> ATA-6 disk at ata0-master
> ad0: 38166MB (78165360 sectors), 77545 C, 16 H, 63 S, 512 B
> ad0: 16 secs/int, 1 depth queue, UDMA100
> GEOM: new disk ad0
> ar: FreeBSD check1 failed
> ata1-master: pio=0x0c wdma=0x22 udma=0x42 cable=40pin
> ata1-master: setting PIO4 on SiS 5513 chip
> ata0-master: setting UDMA33 on SiS 5513 chip
> [0] f:80 typ:165 s(CHS):0/1/1 e(CHS):1023/254/63 s:63 l:78156162
> [1] f:00 typ:0 s(CHS):0/0/0 e(CHS):0/0/0 s:0 l:0
> [2] f:00 typ:0 s(CHS):0/0/0 e(CHS):0/0/0 s:0 l:0
> [3] f:00 typ:0 s(CHS):0/0/0 e(CHS):0/0/0 s:0 l:0
> GEOM: Configure ad0s1, start 32256 length 40015954944 end 40015987199
> GEOM: Configure ad0s1a, start 0 length 268435456 end 268435455
> GEOM: Configure ad0s1b, start 268435456 length 476880896 end 745316351
> GEOM: Configure ad0s1c, start 0 length 40015954944 end 40015954943
> GEOM: Configure ad0s1d, start 745316352 length 3221225472 end 3966541823
> GEOM: Configure ad0s1e, start 3966541824 length 4294967296 end 8261509119
> GEOM: Configure ad0s1f, start 8261509120 length 8589934592 end 16851443711
> GEOM: Configure ad0s1g, start 16851443712 length 23164511232 end 
> 40015954943
> acd0: <HL-DT-ST GCE-8525B/1.03> CDRW drive at ata1 as master
> acd0: read 6890KB/s (8957KB/s) write 8957KB/s (8957KB/s), 2048KB buffer, 
> UDMA 33
> acd0: Reads: CDR, CDRW, CDDA stream, packet
> acd0: Writes: CDR, CDRW, test write, burnproof
> acd0: Audio: play, 255 volume levels
> acd0: Mechanism: ejectable tray, unlocked
> acd0: Medium: CD-R 120mm data disc
> 
> (At this point, the system simply freezes. How would I be able to grab 
> *all* the output from this? Is there an option on this boot CD to enable 
> a serial cable connection?)

Uhm, I miss the vital parts though, but I think I can glean most of what 
I wanted, could you supply me with the output of pciconf -l ? (that can 
be from the 5.2.1 kernel)..

> (At this point, the system simply freezes. How would I be able to grab *all* the output from this? Is there an option on this boot CD to enable a serial cable connection?

Yeah use a serial console, its in the handbook IIRC...

>> Have you tried disabling ACPI and/or APIC ? Both the SiS systems here have had very diffrent success with both of those during times...
>>
> Yes, I have tried disabling both of these in the BIOS without success. The system in question is one of those small-form-factor cubes from iDeq, with the SiS 651 B/962 chipset.

No no, not in the BIOS, that wont change anything, you should boot 
without them (you can choose so in the little menu before the kernel 
boots)..

Also from the old dmesg I see you have atapicam in there, please remove 
that so we can get that out of the equation for the hang...

-- 

-Søren