RFR (XL): 8075492: update IGV in hs-comp

Michael Haupt michael.haupt at oracle.com
Wed Apr 29 15:18:17 UTC 2015

Hi Roland,

> Am 29.04.2015 um 16:45 schrieb Roland Westrelin <roland.westrelin at oracle.com>:
>>>> Webrev: http://cr.openjdk.java.net/~mhaupt/8075492/webrev.00
>>> Why do you need the change to BranchData::print_data_on()?
>> the extra line break otherwise emitted would confuse IGV's input parsing logic, generating a plethora of warning messages.
> BranchData::print_data_on() is also used by PrintMethodData which prints a nicely formatted dump of profiling data. BranchData is not the only profiling data that is printed on multiple lines. Why does this one in particular confuses the IGV?

The dump contains CDATA sections with method bytecode listings. One example is this:

9 if_icmple 17
  16  bci: 9    BranchData          taken(6701) displacement(32)
                                    not taken(0)

Note how the "not taken(0)" appears on an extra line. This extra line confuses the IGV parser, and it will issue a warning like "no match: not taken(0)" for each such case.

The point you're making is a good one; I'll take another look at whether the parsing logic could be improved.




Dr. Michael Haupt | Principal Member of Technical Staff
Phone: +49 331 200 7277 | Fax: +49 331 200 7561
Oracle Java Platform Group | HotSpot Compiler Team 
Oracle Deutschland B.V. & Co. KG, Schiffbauergasse 14 | 14467 Potsdam, Germany
 <http://www.oracle.com/commitment>	Oracle is committed to developing practices and products that help protect the environment

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.openjdk.java.net/pipermail/hotspot-compiler-dev/attachments/20150429/389bb14a/attachment-0001.html>

More information about the hotspot-compiler-dev mailing list