[jdk11u-dev] RFR: 8210988: Improved handling of compiler warnings in the build [v6]
Andrew John Hughes
andrew at openjdk.org
Sat Jul 6 15:10:42 UTC 2024
On Tue, 26 Mar 2024 07:21:54 GMT, Andrew Lu <andrewlu at openjdk.org> wrote:
>> I backport this for parity with 11.0.24-oracle.
>> make/hotspot/lib/CompileGtest.gmk
>> ‘DISABLED_WARNINGS_microsoft := $(DISABLED_WARNINGS_microsoft) \
>> 4146, \’
>> make/hotspot/lib/CompileJvm.gmk
>> DISABLED_WARNINGS_microsoft := $(DISABLED_WARNINGS_microsoft) 4146, \
>>
>> The DISABLED_WARNINGS_microsoft change to 4146 as in src\hotspot\cpu\aarch64\register_aarch64.hpp(285)
>> 'uint32_t first = _bitset & -_bitset' remains.
>
> Andrew Lu has updated the pull request incrementally with one additional commit since the last revision:
>
> update
I found a number of issues with this backport, detailed in [JDK-8335801](https://bugs.openjdk.org/browse/JDK-8335801) and fixed in https://github.com/openjdk/jdk11u-dev/pull/2838
I don't see why this was backported at all. There is no strong reason given in the approval request. Oracle parity is not a sufficient reason alone in most cases, and certainly not for build issues. Oracle have to worry about one build environment (their own) when backporting fixes. As a FOSS project, OpenJDK 11 has to be buildable on a myriad of different compilers, operating systems and architectures. Backports that alter the build system should not be undertaken lightly and they certainly need to be done more carefully than this one was.
-------------
PR Comment: https://git.openjdk.org/jdk11u-dev/pull/2606#issuecomment-2211792915
More information about the jdk-updates-dev
mailing list