Netbeans nb_native needs adjusting to new jdk10 structure

Hi all, It seems to me make/nb_native project is not yet fully adapted after convesion to the new repo structure? When I open the project, I see all native sources greyed out (in addition to dead links to the old sources which can just be deleted). Is anyone working on this? If not, does anyone

Re: Building OpenJDK9 on MSYS2

Hello Peter, On 2017-10-11 00:18, Peter Budai wrote: Thanks Magnus & Erik First thanks for your support and kind words! Magnus, I have checked .bash_profile, .bashrc but they seem to be empty (everything is commented out). You can check with a default MSYS2 install, I have not changed thes

Re: Changing javac default initial heap size

Hello, I believe Wade is rather asking about the resulting javac binary that the build produces. If that's the case, you need to edit jdk/make/launcher/Launcher-jdk.compiler.gmk and add your desired flags to the JAVA_ARGS argument to SetupBuildLauncher for javac. Example: $(eval $(call Setup

RE: Building OpenJDK9 on MSYS2

Hi Erik, The .d file looks like this: C:/msys64/home/peterbud/jdk9/build/windows-x86_64-normal-server-release/hotspot/variant-server/tools/adlc/objs/adlparse.obj: \ C:/msys64/home/peterbud/jdk9/hotspot/src/share/vm/adlc/adlparse.cpp \ I have checked .d.targets file, and looks like it has the fir

Re: Building OpenJDK9 on MSYS2

For gcc, we let the compiler generate the .d file. For the Microsoft tool chain, we use a clever sed script to extract and create it ourself. I think that logic is checking for "Windows", not "Microsoft". That might be your cause of trouble. Look in NativeCompilation.gmk. /Magnus > 11 okt.

RFR: 8189115 - Pass JIB_DATA_DIR to jtreg harness

Please review this small change that allows a user to specify a location for the JIB cache when running jtreg tests that use the JIB Artifact resolver. Webrev: http://cr.openjdk.java.net/~ctornqvi/webrev/8189115/webrev.00/ Thanks

Re: Building OpenJDK9 on MSYS2

Hello, On 2017-10-11 15:48, Magnus Ihse Bursie wrote: For gcc, we let the compiler generate the .d file. For the Microsoft tool chain, we use a clever sed script to extract and create it ourself. I think that logic is checking for "Windows", not "Microsoft". That might be your cause of troubl

Re: RFR: 8189115 - Pass JIB_DATA_DIR to jtreg harness

Looks good. /Erik On 2017-10-11 16:09, Christian Tornqvist wrote: Please review this small change that allows a user to specify a location for the JIB cache when running jtreg tests that use the JIB Artifact resolver. Webrev: http://cr.openjdk.java.net/~ctornqvi/webrev/8189115/webrev.00/

RFR: JDK-8189095; Import JMC from artifactory using Jib and main makefiles

Please review this small fix for the mac-bundles target. The current solution does not handle files with spaces in them. By changing this to a single rule with a recursive copy, any filenames that includes spaces will be handled correctly. Webrev: http://cr.openjdk.java.net/~erikj/8189095/webr

Re: RFR: JDK-8189095; Import JMC from artifactory using Jib and main makefiles

Erik: Please review this small fix for the mac-bundles target. The current solution does not handle files with spaces in them. By changing this to a single rule with a recursive copy, any filenames that includes spaces will be handled correctly. Webrev: http://cr.openjdk.java.net/~erikj/8189095

Re: Changing javac default initial heap size

Thanks, that's exactly what I was looking for. -wade On Wed, Oct 11, 2017 at 3:21 AM, Erik Joelsson wrote: > Hello, > > I believe Wade is rather asking about the resulting javac binary that the > build produces. If that's the case, you need to edit > jdk/make/launcher/Launcher-jdk.compiler.gmk a