JDK-8248901: Signed immediate support in [...] breaks 8u build on Windows/RHEL 6

[Severin Gehwolf]

Hi Andrew,

On Tue, 2021-08-10 at 06:26 +0100, Andrew Hughes wrote:
> On 18:22 Mon 09 Aug     , Severin Gehwolf wrote:
> > Hi,
> > 
> > JDK-8248901 got backported to 8u and is included in 8u312-b01. It ended
> > up there as a pre-requisite for JDK-8247766 which isn't yet approved
> > (thus, pushed). This breaks the Vanilla JDK 8u builds for Linux (built
> > on RHEL/CentOS 6) and Windows.
> > 
> > See:
> > https://bugs.openjdk.java.net/browse/JDK-8270379?focusedCommentId=14440384&page=com.atlassian.jira.plugin.system.issuetabpanels%3Acomment-tabpanel#comment-14440384
> > 
> > and
> > 
> > https://bugs.openjdk.java.net/browse/JDK-8270379?focusedCommentId=14440626&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-14440626
> > 
> > Could somebody please look into it or shall we rather revert?
> > 
> > Opinions?
> > 
> > Thanks,
> > Severin
> > 
> 
> On RHEL 6, the definition of {U,}INT64_C is guarded by
> __STDC_CONSTANT_MACROS. If we define that in globalDefinitions.hpp,
> or globalDefinitions_gcc.hpp, the build works on that platform.
> 
> Comparing stdint.h on RHEL 6 with the one on RHEL 8 shows:
> 
> -/* The ISO C99 standard specifies that in C++ implementations these
> -   should only be defined if explicitly requested.  */
> -#if !defined __cplusplus || defined __STDC_CONSTANT_MACROS
> -
> 
> In the webrev [0], I've added it to globalDefinitions.hpp in the hope
> it may also work on Windows. Please give this a try.

Thanks, testing it now.

> Can you also file a bug for this, rather than commenting on the
> backport bug? We'll need a bug ID even to do a full reversion.

Here you go:
https://bugs.openjdk.java.net/browse/JDK-8272214

Thanks,
Severin

> [0] https://cr.openjdk.java.net/~andrew/openjdk8/8248901-fail/webrev.01/
> 
> Thanks,