ksh93 is a POSIX conform shell, this includes C99 conform arithmetic expressions in shell arithmetic expressions, i.e. (( expression )). The C99 standard and POSIX require that NaN has a sign (and a payload, but this is not this topic). This is not a bug, this is an intentional feature which becomes useful to detect from which side (positive, negative or above, below) the NaN was approached (e.g. copysign() should still work on NaN values).
Cleaning the sign bit would be a bug and the test suite would choke on it, there is even an explicit test in the shell test suite which verifies whether negative NaN is functional. I don't know whether the difference between SPARC and i386 is a bug in i386 or the standard undefined. This should be investigated by the compiler team. Olga On Mon, Apr 12, 2010 at 3:45 PM, Keith Bierman <[email protected]> wrote: > Any reason the Solaris ksh port doesn't clear the sign bit before calling > libc? I presume that changing the ksh source globally would take the authors > cooperation, but would be ideal > > [email protected] | [email protected] > Sent from my iPod > On Apr 12, 2010, at 12:52 AM, "Zhishun A. Liu" <[email protected]> wrote: > > Please see my posting at > http://opensolaris.org/jive/thread.jspa?threadID=127167#473647 > > Bottom line: Intel hardware isn't broken; NaN *display* has nothing to do > with the compiler used. > -- > This message posted from opensolaris.org > _______________________________________________ > tools-compilers mailing list > [email protected] > > ________________________________ > The information contained in this transmission may be confidential. Any > disclosure, copying, or further distribution of confidential information is > not permitted unless such privilege is explicitly granted in writing by > Quantum. Quantum reserves the right to have electronic communications, > including email and attachments, sent across its networks filtered through > anti virus and spam software programs and retain such messages in order to > comply with applicable data security and retention requirements. Quantum is > not responsible for the proper and complete transmission of the substance of > this communication or for any delay in its receipt. > > _______________________________________________ > tools-compilers mailing list > [email protected] > -- , _ _ , { \/`o;====- Olga Kryzhanovska -====;o`\/ } .----'-/`-/ [email protected] \-`\-'----. `'-..-| / Solaris/BSD//C/C++ programmer \ |-..-'` /\/\ /\/\ `--` `--` _______________________________________________ tools-compilers mailing list [email protected]
