|
|
CT60 and related things BBS
| Re: CTPCI new CPLD |
|---|
|
Posted by: Rodolphe
|
Dec,26.2011-12:22
|
Do not use the 1G CPLD, it has yet a bug in the arbiter.
Should be solved this day.
I can see some PERR now; very rare beacause the most parity error are not on the data phase but on the address phase (begining with FRAME) and this concludes with a TARGET ABORT from PLX (thjat is the target of the NEC)... This is a normal behaviour for a TARGET seeing bad parity on a add phase.
The SERR signal should output.
The question was (this night) WHY the target abort ?
And I think I found : a GNT timing problem and a collision with NEC driving the bus (with ADD phase) and the PLX triyng a new transaction and driving too the bus...and disturbing the add + c/be bits from the NEC ---> parity error on add phase ---> target abort from PLX. And some time Target abort on NEC too because the PLX add phase is parity error too !
Seems now an evidence.
|
[All messages in this thread] [Start new thread]
What's the anti-troll code?
That's your personal code to be able to add comments and messages on the dhs.nu site.
Don't have a code or forgot it? Fix it here.
|
