[prev in list] [next in list] [prev in thread] [next in thread] 

List:       ltp-coverage
Subject:    Re: [Ltp-coverage] Improving --demangle-cpp performance
From:       Peter Oberparleiter <oberpar () linux ! vnet ! ibm ! com>
Date:       2012-11-29 12:41:50
Message-ID: 50B7580E.7090801 () linux ! vnet ! ibm ! com
[Download RAW message or body]

Hi,

On 23.11.2012 07:23, Olly Betts wrote:
> Currently --demangle-cpp is very slow because it runs c++filt each time
> there's a symbol to decode.  For example, for Xapian (http://xapian.org/)
> c++filt gets run 14326 times, and passing --demangle-cpp makes genhtml
> take almost 5 times as long:

Thanks for the thorough analysis. I agree that this should be improved 
and your approach of using c++filt in filter mode sounds promising. Your 
current patch leaves some error recovery questions open though:

1. According to the man page, open3 doesn't return false if the 
specified command fails
2. Should the c++filt process die somewhere along the way, the next 
write/read would cause a SIGPIPE which would need to be handled accordingly.

I'll see if I can add this handling.


Regards,
   Peter Oberparleiter

-- 
Peter Oberparleiter
Linux on System z Development - IBM Germany


------------------------------------------------------------------------------
Keep yourself connected to Go Parallel: 
VERIFY Test and improve your parallel project with help from experts 
and peers. http://goparallel.sourceforge.net
_______________________________________________
Ltp-coverage mailing list
Ltp-coverage@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/ltp-coverage
[prev in list] [next in list] [prev in thread] [next in thread]