Re: [partial-reconfig] ICAP on Virtex II Pro and ML310

[<il-guru@xxxxxxxxxxx>]

oh sorry i didn't explain:
on write-word status i send to icap_in port 1 byte @ time.
so 32bit word is divided in 4 byte words.sorry

i'm sure to not violating abort sequence (i change WR only 
on CE=1).

i've tried to wait for 16k clock cycles...but my BUSY 
doesn't change.

you said "all signals are active-low". ICAP_IN too?

thanks.
Davide


On Wed, 7 May 2008 02:07:36 +0300
 "Mati Nahshon" <mati.nahshon@xxxxxxxxx> wrote:
> Hi,
> It's been long time since I worked on it but just some 
> points that may help:
> - all signals are active low
> - Big endian bytewize = to write AA995566 do:
> AA
> 99
> 55
> 66
> - not every thing is documented try adding 20000000 if 
> things don't work.
> - Abort status word is presented on ICAP_O[4:7] while 
> read is not performed
> (this is not documented) this can help during debug 
> (look for errors are
> status during operation)
> - make sure you are not violating the abort rule (double 
> check).
> - are you waiting long enough?
>
> Mati
>
> -----Original Message-----
> From: owner-partial-reconfig@xxxxxxxxxxxxxx
> [mailto:owner-partial-reconfig@xxxxxxxxxxxxxx] On Behalf 
> Of Bollo Davide
> Sent: Wednesday, May 07, 2008 12:40 AM
> To: partial-reconfig@xxxxxxxxxxxxxx
> Subject: [partial-reconfig] ICAP on Virtex II Pro and 
> ML310
>
> Hi all, i'm newbie!
>
> I'm working on a State Machine to command directly ICAP 
> interface on a
> Virtex II Pro XCV2P30 on a ML310 board.
>
> I'd followed instruction on UG012, User Guide by Xilinx 
> about SELECT MAP
> communication Protocol. I know that Select Map DATAPORT 
> have "0" as MSB, so
> ICAP have "7" as MSB.
>
> I made a State Machine in VHDL code, to command ICAP 
> interface, and read
> IDCODE from my device, with these step:
>
>        CE    WR        WORD
> 1        1        0
>
> 2        0        0        FFFFFFFF		Dummy word
> 3        0        0        AA995566		synch
> 4        0        0        2801C001		read 1 word from 
> IDCODE
> 5        0        0        20000000		flush word
> 6        0        0        20000000
>
> 7        1        0                       (here busy 
> goes HIGH)
> 8        1        1
> 9        0        1                       waiting for 
> busy=0 --> read word
> (4 CK) --> go to the next state
>
> 10      1        1
> 11      1        0
> 12      0        0
>
> 13      0        0        0000000D    desynch
> 14      0        0        20000000
> 15      0        0        20000000
>
> Every write-word step is splitted in 4 substeps.
> Machine works good on simulation (with busy simulation 
> too).
>
> I snoop signal with Chipscope, with ILA core insertion 
> on every icap signal.
>
>
> My machine goes on step 9 and wait here forever.
> Why BUSY doesn't return down?
> Where is my mistake?
>
> Some information:
> 	Icap_clock= 10 Mhz
> 	Ise 7.1, Chipscope 7.1 (old but i can't work only with 
> this)
> 	download bitstream configuration on JTAG chain, with 
> chipscope.
> 	M0:M2 = 111
>
> I hope in your help!
>
>
> Davide
> 	
>
>
> ___________________________
> partial-reconfig mailing list
> partial-reconfig@xxxxxxxxxxxxxx
> Mailing List Archive : 
> http://www.itee.uq.edu.au/~listarch/partial-reconfig/
>
> ___________________________
> partial-reconfig mailing list
> partial-reconfig@xxxxxxxxxxxxxx
> Mailing List Archive : 
> http://www.itee.uq.edu.au/~listarch/partial-reconfig/


--------------------------------------------------------------------------
Hai tempo fino al 31/03/2008 per azzerare i costi di attivazione:
da casa o dall'ufficio naviga con la qualità di Infinito ADSL
(www.infinito.it/adsl)
___________________________
partial-reconfig mailing list
partial-reconfig@xxxxxxxxxxxxxx
Mailing List Archive : http://www.itee.uq.edu.au/~listarch/partial-reconfig/