[Asterisk-Dev] Zaptel/libpri/te110p guru help needed (Solaris port)

[Simon Lockhart]

Okay, so I'm in the middle of porting zaptel et al to Solaris, and whilst I've
had some success, I've hit a couple of problems which I'm hoping someone who
has spent a bit longer than me in the guts of zaptel will be able to help
me with.

[Load module]
Dec 23 20:25:04 dev1.slimey.org zaptel: Registered Span 1 ('WCT1/0') with 31 channels
Dec 23 20:25:04 dev1.slimey.org zaptel: Span ('WCT1/0') is new master
Dec 23 20:25:04 dev1.slimey.org wcte11xp: Found a Wildcard: Digium Wildcard TE110P T1/E1

[run ztcfg]
Dec 23 20:31:25 dev1.slimey.org wcte11xp: TE110P: Span configured for CCS/HDB3/CRC4
Dec 23 20:31:25 dev1.slimey.org wcte11xp: Calling startup (flags is 4099)

dev1.slimey.org-root-10: modinfo | grep -i zap
237  131976c  335f3 244   1  zaptel (Zapata Telephony Interface)
256  1387b6e   6e26 243   1  wcte11xp (Wildcard T100P/E100P Zaptel Dri)

All good so far...

I've got the TE110P connected back-to-back with another TE110P in a linux box,
which is a 'known working' configuration.

Starting asterisk on the Solaris box, I get a whole load of the following 
errors:

Dec 23 20:39:31 WARNING[384]: chan_zap.c:7257 zt_pri_error: PRI: Short write: -1/6 (Error 0)

(although, I'm guessing these will go away when I fix some of the other 
problems)

*CLI> pri show span 1
Primary D-channel: 16
Status: Provisioned, Up, Active
Switchtype: EuroISDN
Type: CPE
[snip]

So, the E1 is coming up, so I'm assuming it's talking fine to the card 
(asterisk talking to libpri, talking to zaptel, talking to wcte11xp module.
Module talking to the TE110P card, so DMA working fine).

However...

Until I commented it out of the code in chan_zap.c, I'm getting this message
continuously:

Dec 22 22:43:38 WARNING[321]: Whoa....  I'm reading but have no cidspill (15)...

FD 15 is the first (and only configured) BChan on the E1. I don't see this on
the same config on the linux box, so I'm assuming it's a problem with the
solaris port.

I'm assuming this is caused by data appearing on a BChan which isn't active
yet. Is this actually the case, or are all the BChans initialised at startup?
Where should I start looking to try and fix this? Asterisk is taking up almost
100% of CPU, and I'm assuming this is because it's in a continuous loop trying
to read data (poll says there's data to read), but because there's no cidspill
it never gets as far as trying to read it, so the next time it polls, it tries
again.

I'm also seeing this occasionally:

Dec 23 11:39:57 dev1.slimey.org wcte11xp: Expecting top ca1e, got d5d5

but I'm guessing this is because it's not managing to service all the 
interrupts because the CPU is maxed.

Any pointers?

And, while I've got the attention of zaptel experts, I'm having more serious
problems trying to get X100P and TDM400P working under solaris. Using 
essentially the same method of porting, I get random bus errors from the kernel
when doing inb/outb to the card (or rather the solaris equivalent). Now, I
know that all three cards (X100P, TDM400P and TE110P) all use the same PCI
interface chip (Tigerjet 320), so I'm wondering if there are any subtle
differences in the hardware which could cause the TE110P to not exhibit the
same problems...

Thanks,

Simon
-- 
Simon Lockhart | * Sun Server Colocation * ADSL * Domain Registration *
   Director    |    * Domain & Web Hosting * Internet Consultancy * 
  Bogons Ltd   | * http://www.bogons.net/  *  Email: info at bogons.net  *