--- Quote Start ---
originally posted by slacker+dec 1 2006, 04:13 pm--><div class='quotetop'>quote (slacker @ dec 1 2006, 04:13 pm)</div>
--- quote start ---
<!--quotebegin-karel van haver@Dec 1 2006, 12:07 PM
dear all,
i have a problem that is maybe already seen by one of you:
i have generated a nios core with a 256kx16 bit sram interface. the sram module is connected to the avalon tri-state bridge.
address = (0-17)
data = (0-15)
wrn = 1 bit
rdn = 1 bit
csn = 1 bit
bhen=1 bit
blen = 1 bit
the generation of the nios core completes without errors. but after i have finished my design fitting .. i get following warning
sram_address(0) is synthezised away
any idea why?... is the address mapping between the avalon bridge and the sram module wrong?
regards
<div align='right'><{post_snapback}> (index.php?act=findpost&pid=19768)
--- quote end ---
--- Quote End ---
Read the building memory subsystems (
http://www.altera.com/literature/hb/qts/qts_qii54006.pdf) section in the SOPC Builder handbook. Look at table 9-1... You're welcome to either modify the existing Europa-based (Perl HDL generation) SRAM component or generate your own using the Component Editor/Builder.
Best Regards,
- slacker
<div align='right'><{post_snapback}> (index.php?act=findpost&pid=19769)</div>
[/b]
--- Quote End ---
Slacker,
Thanks a lot for the spec sheet. So, my thought that the mapping between the SRAM address and the tri state address pinning wasn't correct and is indeed explained in the spec sheet.
So, instead of connecting SRAM A0 pin to tri-state-bridge-address0, I should connect SRAM A0 to tri-state-bridge-address1 and keep tri-state-bridge-address0 unconnected.
Regards
Karel