Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Synthesis on Xilinx ISE 10.1 hang on for a long time

Status
Not open for further replies.

bamboolll

Newbie level 2
Joined
Jul 23, 2010
Messages
2
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,281
Location
HCM city
Activity points
1,298
I had a project which i must implement Regular Expression Matching on FPGA. I had chosen Nondeterminate Final Automat (NFA), a kind of Finite State Machine (FSM) to implement it.
Because of using FSM, so in my project there are so many states to implement a list of Regular Expression. With the list of 20 Regular Expression, it take about 1100 State, at this level XST on ISE synthesis just fine, about 3-4 minute to do. But with the list of 100 Regular Expression, XST hang on for a long time. I didn't see any sign which indicate that it still working.
Anyone can give me some advice for this situation?
Can i separate synthesis with XST ?
is there any way to XST show the log when it synthesize. I want to know what make XST hang on for a long time.
 

XST does have some issues. One issue happens when XST runs out of memory. As with any program where heavy paging occurs, the slowdown is extreme.

this should be easy to spot using "top" in linux, or looking at the resources (ctrl+alt+del, then select the correct option) in windows. XST will generally use 90-100% of 1 CPU, but shouldn't use 100% of the RAM.
 

Thank for your reply. The problem of running out of memory is true. When i used ISE 8.1 for synthesis, the using memory increase without stop, until it ran out of memory (i had 3GB DDRAM2).
When i change to ISE 10.1. The memory cost when synthesis process didn't raise so much, but CPU cost it still at 100%. That why after 40 minitues of waiting, i must stop ISE by end process command.
Do you think it is a problem of FSM?
 

i have no idea how you wrote the code, but its possible that either some optimization has gotten stuck in a loop, or maybe something in your code did. or possibly you've hit some case that is just very slow.

you can always look at the .syr file during synthesis to see what is changing. At least I think you can -- windows might not allow it but linux does.

Its also possible that synthesis would finish after a few hours.
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top