Re: sparc implementation

From: Vincentius Robby (vincentius_at_umich_dot_edu)
Date: Fri Sep 05 2008 - 11:47:01 PDT

  • Next message: Hideyuki Jitsumoto: "UTS_VERSION unmatched"
    Hello Paul,
    
    Yes, those corrections eliminated the warnings. Only the unresolved  
    symbols errors are left now. Where am I actually supposed to look for  
    these? I tried blindly grepping for the symbol name and found nothing.
    
    Thanks.
    
    -- 
    Vincentius Robby
    
    
    Quoting "Paul H. Hargrove" <PHHargrove_at_lbl_dot_gov>:
    
    > I see 2 problems.
    >
    > 1) cri_syscall3() has "CRzI_SYSCALL_STRING", where clearly the "z" is a typo.
    >
    > 2) In all the syscall macros we have
    > 	register long __g1 __asm__ ("g1") = name;
    > but should, instead have,
    > 	register long __g1 __asm__ ("g1") = nr;
    > This is the one causing the warning.
    >
    > This "name" came from the glibc macros, but we use "nr" and have  
    > used "name" elsewhere in the same macro (for "cri_open", for  
    > instatance).
    >
    > -Paul
    >
    > -- 
    > Paul H. Hargrove                          PHHargrove_at_lbl_dot_gov
    > Future Technologies Group
    > HPC Research Department                   Tel: +1-510-495-2352
    > Lawrence Berkeley National Laboratory     Fax: +1-510-486-6900
    >
    >
    >
    

  • Next message: Hideyuki Jitsumoto: "UTS_VERSION unmatched"