Re: RFR: 8247396: PCH should be disabled for opt-size feature with clang after JDK-8246751

2020-06-12 Thread Magnus Ihse Bursie
On 2020-06-12 04:58, jiefu(傅杰) wrote: Hi Erik and Magnus, Thanks for your help. Updated: http://cr.openjdk.java.net/~jiefu/8247396/webrev.02/ Also I've changed the JBS title to 'Exclude PCH for files in $(OPT_SPEED_SRC) with clang after JDK-8246751'. Please review it and give me some adviece.

Re: 8247396: PCH should be disabled for opt-size feature with clang after JDK-8246751(Internet mail)

2020-06-12 Thread 傅杰
Thanks Magnus for your review and help. I can push it myself and will do it later. Best regards, Jie On 2020/6/12, 5:08 PM, "Magnus Ihse Bursie" wrote: On 2020-06-12 04:58, jiefu(傅杰) wrote: > Hi Erik and Magnus, > > Thanks for your help. > > Updated: http://cr.openjdk.

RFR: JDK-8150828 Consider using '-fno-asynchronous-unwind-tables' to reduce the size of libjvm.so by 10 percent

2020-06-12 Thread Magnus Ihse Bursie
From Volker's bug report: "We are building and linking the libjvm.so on Linux with -fnoexceptions because we currently don't use C++ exception handling in the HotSpot. Nevertheless, g++ generates unwind tables (i.e. .eh_frame sections) in the object files and shared libraries which can not be

Re: RFR: JDK-8150828 Consider using '-fno-asynchronous-unwind-tables' to reduce the size of libjvm.so by 10 percent

2020-06-12 Thread Erik Joelsson
Looks good to me at least. /Erik On 2020-06-12 05:21, Magnus Ihse Bursie wrote: From Volker's bug report: "We are building and linking the libjvm.so on Linux with -fnoexceptions because we currently don't use C++ exception handling in the HotSpot. Nevertheless, g++ generates unwind tables

RFR: JDK-8200738 Use --hash-style=gnu for gcc

2020-06-12 Thread Magnus Ihse Bursie
When support for gnu hash tables were added to the gcc toolchains, the OpenJDK build system added a hack to change --hash-style=gnu to --hash-style=both unconditionally, citing compatibility concerns. --hash-style=gnu has a very slight performance advantage over --hash-style=both (very small s

Re: RFR: JDK-8200738 Use --hash-style=gnu for gcc

2020-06-12 Thread Erik Joelsson
Looks good. /Erik On 2020-06-12 08:42, Magnus Ihse Bursie wrote: When support for gnu hash tables were added to the gcc toolchains, the OpenJDK build system added a hack to change --hash-style=gnu to --hash-style=both unconditionally, citing compatibility concerns. --hash-style=gnu has a ver

Re: RFR(S) : 8211977 : move testlibrary tests into one place

2020-06-12 Thread Igor Ignatyev
adding build-dev > On Jun 12, 2020, at 8:36 PM, Igor Ignatyev wrote: > > http://cr.openjdk.java.net/~iignatyev//8211977/webrev.00/ >> 796 lines changed: 200 ins; 588 del; 8 mod; > > Hi all, > > could you please review this small patch which puts all tests for testlibrary > classes into one lo

Re: RFR(S) : 8211977 : move testlibrary tests into one place

2020-06-12 Thread Igor Ignatyev
testing revealed that LingeredAppTest.java required some love, incremental webrev w/ the fixes for LingeredAppTest -- http://cr.openjdk.java.net/~iignatyev//8211977/webrev.0-1 -- Igor > On Jun 12, 2020, at 8:38 PM, Igor Ignatyev wrote: > > adding build-dev > >> On Jun 12, 2020, at 8:36 PM, I