Re: C4819 warnings on Windows

2019-01-03 Thread David Holmes
Hi Yasumasa, The two hotspot issues can be filed in one bug under hotspot->compiler and reviewed on hotspot-compiler-dev as they are source code changes not build changes. The other one is a build change but probably also needs reviewing by awt-dev. Cheers, David On 4/01/2019 5:22 pm, Yas

C4819 warnings on Windows

2019-01-03 Thread Yasumasa Suenaga
Hi all, I tried to build OpenJDK on WSL (Windows 10 1809 + VS2017 (15.9.4) + Ubuntu 18.04 LTS). However, I saw some C4819 warnings as below: ``` c:/OpenJDK/jdk/src/hotspot/share/compiler/methodMatcher.cpp(258): warning C4819: ファイルは、現在のコード ページ (0) で表示できない文字を含んでいます。データの損失を防ぐために、ファイルを Unicode 形式

Re: RFR: JDK-8216021: RunTest.gmk might set concurrency level to 1 on Windows

2019-01-03 Thread Tim Bell
Erik: When running tests using "make run-test-prebuilt" on Windows, the calculation of available memory may fail. This only happens on certain systems, and is likely related to which version of Cygwin that is being used. The consequence is that concurrency gets set to 1. The cause of the problem

Re: RFR: JDK-8216021: RunTest.gmk might set concurrency level to 1 on Windows

2019-01-03 Thread Christian Tornqvist
Looks good, thanks for fixing this! Thanks, Christian > On Jan 3, 2019, at 2:30 AM, Erik Joelsson wrote: > > When running tests using "make run-test-prebuilt" on Windows, the calculation > of available memory may fail. This only happens on certain systems, and is > likely related to which ver

RE: JDK-8216048: Fix devkit creation in WSL

2019-01-03 Thread Andrew Luo
Hi Erik, Thanks for sponsoring this. Looking at the patch, I noticed I left in an extra "echo "$VS_INSTALL_DIR/$f"". This probably could be removed. Thanks, -Andrew -Original Message- From: Erik Joelsson Sent: Thursday, January 3, 2019 1:22 AM To: Andrew Luo ; build-dev@openjdk.jav

RFR: JDK-8216021: RunTest.gmk might set concurrency level to 1 on Windows

2019-01-03 Thread Erik Joelsson
When running tests using "make run-test-prebuilt" on Windows, the calculation of available memory may fail. This only happens on certain systems, and is likely related to which version of Cygwin that is being used. The consequence is that concurrency gets set to 1. The cause of the problem is m

RFR: JDK-8216048: Fix devkit creation in WSL

2019-01-03 Thread Erik Joelsson
Hello Andrew, Thanks for fixing this. The patch looks good. I have verified that it produces the same output in WSL and Cygwin. Just like in Cygwin, I did not need to be in a Tools Command Prompt to run it (because VS was installed in the default location). Regarding the hard coded dir versi

Re: RFR: JDK-8215952: NetBeans pre-build failed due to the incorrect configuration

2019-01-03 Thread Fu Jie
Hi Alan, With NetBeans, we can edit all the sources in the JDK just like using a vim. But I don't think that is the main intent of the NB project. For me, I found it more convenient to analyze the source code such as finding the definition and usages of a method, showing the call graph in a me

Re: RFR: JDK-8215952: NetBeans pre-build failed due to the incorrect configuration

2019-01-03 Thread Alan Bateman
On 03/01/2019 07:22, Fu Jie wrote: Thanks Alan for your review and valuable advice. I think it's worth keeping a NB project in the repo since it seems more convenient for IDE developers. And to keep configurations.xml current is also important. I made a patch to fix the issue mentioned by Ala