bluespec.com Forum Index bluespec.com
Bluespec Forums
 
 FAQFAQ   SearchSearch   MemberlistMemberlist   UsergroupsUsergroups   RegisterRegister 
 ProfileProfile   Log in to check your private messagesLog in to check your private messages   Log inLog in 

Compilation stuck in the transform stage.

 
Post new topic   Reply to topic    bluespec.com Forum Index -> Tools: BSC (Bluespec Compiler)
View previous topic :: View next topic  
Author Message
akshayj



Joined: 19 Feb 2015
Posts: 10

PostPosted: Fri Sep 18, 2015 6:40 am    Post subject: Compilation stuck in the transform stage. Reply with quote

Hey,

I have a code in which PulseWires and RWires are used to get the result within one clock cycle.
If the Vector of PulseWire type is [16] x [16], the code is compiled successfully and gives desired result.

Code:
        Vector#(16,Vector#(16,PulseWire)) pulsewire     <- replicateM(replicateM(mkPulseWire)); //Pulsewires to find minimum


But when I increase the Vector for ex: [64] x [64], the compilation gets stuck in the Transform satge (used -v flag).

What may be the reason behind it?

Thanks in advance!!
Akshay
Back to top
View user's profile Send private message
quark
Site Admin


Joined: 02 Nov 2007
Posts: 496

PostPosted: Mon Oct 05, 2015 9:57 am    Post subject: Re: Compilation stuck in the transform stage. Reply with quote

The size of the instantiation is not directly an issue. It's probably the style in which your rule (or rules) are written. There are some known inefficiency issues in the "transform" stage that we're working on. I don't know if that's what you're encountering. If you can show the code, I might be able to offer advice for how to work around it (usually the vector indexing is where the problem occurs). As an experiment, you can also try the internal flag -no-opt-AndOr and see if that improves the compile time. This turns off a boolean optimization that "transform" does. If that's where the inefficiency is happening for you, then that could work around the problem.
Back to top
View user's profile Send private message
akshayj



Joined: 19 Feb 2015
Posts: 10

PostPosted: Mon Oct 05, 2015 2:49 pm    Post subject: Reply with quote

First of all Thanks for the reply. Smile

I tried -no-opt-AndOr flag, but still there was no improvement in the compilation time.

I intend to find the index of the minimum value in an input vector within one clock cycle. Please find the code as an attachment.
I have tried to code it in two ways, but both result in almost same compilation time.
As the size of Input vector is increased to say 64, the compilation time increases to few days. Sad



disparity_calc_modified.bsv
 Description:

Download
 Filename:  disparity_calc_modified.bsv
 Filesize:  4.38 KB
 Downloaded:  468 Time(s)

Back to top
View user's profile Send private message
akshayj



Joined: 19 Feb 2015
Posts: 10

PostPosted: Tue Oct 06, 2015 2:58 am    Post subject: Reply with quote

I noticed that I was using the -O flag, once I removed this flag the compilation was quick.

Now, will removing the optimization flag also result in an unoptimized verilog code/synthesized circuit?

Also is there a work around so that I can use optimization and also bring down the compilation time.

Thanks!
Akshay
Back to top
View user's profile Send private message
Display posts from previous:   
Post new topic   Reply to topic    bluespec.com Forum Index -> Tools: BSC (Bluespec Compiler) All times are GMT - 4 Hours
Page 1 of 1

 
Jump to:  
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
You can attach files in this forum
You can download files in this forum
bluespec.com topic RSS feed 


Powered by phpBB © 2001, 2005 phpBB Group
Protected by Anti-Spam ACP