diff options
Diffstat (limited to 'README.sun386i')
| -rw-r--r-- | README.sun386i | 45 |
1 files changed, 45 insertions, 0 deletions
diff --git a/README.sun386i b/README.sun386i new file mode 100644 index 00000000..e9daf26c --- /dev/null +++ b/README.sun386i @@ -0,0 +1,45 @@ +Date: Mon, 16 Mar 1992 14:49:10 -0400 +From: <beebe@math.utah.edu> +To: david@cs.dal.ca, arnold@skeeve.atl.ga.us + +... + +On the Sun 386i, floating-point numbers are printed without a leading +zero digit. This causes the cmp step to fail for the chem target. I +revised the Makefile, and at the same time, inserted an RM macro in +place of the many rm calls. The complete updated Makefile is appended +below. This change should be applied to all systems to avoid the +leading-zero problem in the future. + +On the Sun 386i, my initial "make sunos40" resulted in a load failure +with + +>> Undefined: +>> strncasecmp +>> strftime + +I therefore modified the Makefile to read + +OPTIMIZE= -g -O -fstrength-reduce -DSTRFTIME_MISSING -DSTRCASE_MISSING + +and loading got further, but still failed: + +>> tzset: ld: /lib/libc.a(localtime.o): multiply defined +>> *** Error code 1 +>> make: Fatal error: Command failed for target `gawk' +>> Current working directory /home/share/gnu/src/gawk-2.13 +>> *** Error code 1 +>> make: Fatal error: Command failed for target `test' + +Investigation with nm showed that localtime() is only referenced by +builtin.o, so I simply did + + chmod +x gawk + make test bigtest + +and was able to complete the test after the leading-zero digit fix +was applied to the Makefile. + +The Sun 386i is not receiving further development by Sun, and SunOS +4.0.3 is the last O/S release for it, so perhaps you don't want to do +anything other than note the problem in the README file. |