Hi Phil, thanks for the review. I'll update the bug and bug and push the change to client as requested.
You're right that the problem doesn't currently manifest in a plain OpenJDK build. But in our proprietary builds we have some additional instrumentation and logging code which includes ctype_iso.h via wchar.h Regards, Volker On Tue, Apr 29, 2014 at 8:00 PM, Phil Race <[email protected]> wrote: > I don't see any harm in this change but I also don't see any warnings about > this > in the RE build log files for Solaris or Linux .. please push to 9/client > (I note your > change set is against 9/dev). Also label the bug 'noreg-cleanup' else SQE > will be asking for a regression test ... > > thanks, > > -phil. > > > On 4/29/2014 10:06 AM, Volker Simonis wrote: >> >> Hi, >> >> could you please review this tiny fix which merely renames some >> variable names to make them C/C++ standard conformant and prevent a >> name clash with a macro defined in in /usr/include/iso/ctype_iso.h on >> Solaris: >> >> http://cr.openjdk.java.net/~simonis/webrevs/8042090/ >> https://bugs.openjdk.java.net/browse/JDK-8042090 >> >> C/C++ forbid variable names which begin with an underscore followed by >> a capital letter. >> >> The C++ Standard [ISO/IEC 14882-2003], Section 17.4.3.1.2 specifies >> the following sets of reserved names: "Each name that contains a >> double underscore __ or begins with an underscore followed by an >> uppercase letter is reserved to the implementation for any use." >> >> The C99 Standard [ISO/IEC 9899:1999] specifies in section 7.1.3 the >> following sets of reserved names: "All identifiers that begin with an >> underscore and either an uppercase letter or another underscore are >> always reserved for any use." >> >> sun/java2d/loops/ProcessPath.c defines the variable names "_X" and >> "_X" where "_X" clashes with the preprocessor macro: >> >> #define _X 0x00000080 /* heXadecimal digit */ >> >> defined in /usr/include/iso/ctype_iso.h on Solaris. >> >> Thank you and best regards, >> Volker > >
