| 1 | Solaris Problem #1:
|
|---|
| 2 | ===================
|
|---|
| 3 | From: carson@lehman.com
|
|---|
| 4 | Date: Fri, 7 Feb 1997 01:05:58 -0500
|
|---|
| 5 | To: arnold@gnu.ai.mit.edu
|
|---|
| 6 | Subject: Solaris 2.5.1 x86 bug in gawk-3.0.2
|
|---|
| 7 |
|
|---|
| 8 | awktab.c has the following bogus logic:
|
|---|
| 9 |
|
|---|
| 10 | #ifndef alloca
|
|---|
| 11 | #ifdef __GNUC__
|
|---|
| 12 | #define alloca __builtin_alloca
|
|---|
| 13 | #else /* not GNU C. */
|
|---|
| 14 | #if (!defined (__STDC__) && defined (sparc)) || defined (__sparc__) || defined (__sparc) || defined (__sgi)
|
|---|
| 15 | #include <alloca.h>
|
|---|
| 16 | #else /* not sparc */
|
|---|
| 17 |
|
|---|
| 18 | Solaris x86 obviously dosn't define sparc or __sparc.
|
|---|
| 19 |
|
|---|
| 20 | What you _meant_ to say was:
|
|---|
| 21 |
|
|---|
| 22 | if (defined(__sun) && defined(__SVR4))
|
|---|
| 23 |
|
|---|
| 24 | (which identifies Solaris 2.x under both Sun's cc and gcc)
|
|---|
| 25 |
|
|---|
| 26 | --
|
|---|
| 27 | Carson Gaspar -- carson@cs.columbia.edu carson@lehman.com
|
|---|
| 28 | http://www.cs.columbia.edu/~carson/home.html
|
|---|
| 29 | <This is the boring business .sig - no outre sayings here>
|
|---|
| 30 |
|
|---|
| 31 | * * * * * * *
|
|---|
| 32 |
|
|---|
| 33 | Solution to Problem #1:
|
|---|
| 34 | =======================
|
|---|
| 35 | Tue Oct 20 21:25:11 IST 1998
|
|---|
| 36 |
|
|---|
| 37 | This has been fixed in 3.1.0 with the bisonfix.sed script.
|
|---|
| 38 |
|
|---|
| 39 | Arnold Robbins
|
|---|
| 40 | arnold@gnu.org
|
|---|
| 41 |
|
|---|
| 42 | Solaris Problem #2:
|
|---|
| 43 | ===================
|
|---|
| 44 | Tue Apr 13 16:57:45 IDT 1999
|
|---|
| 45 |
|
|---|
| 46 | There is a known problem in that the `manyfiles' test will fail under
|
|---|
| 47 | Solaris if you set your soft limit on the number of file descriptors to
|
|---|
| 48 | above 256. This is due to a "feature" of fdopen that an fd must be
|
|---|
| 49 | less than 256 (see fdopen(3)).
|
|---|
| 50 |
|
|---|
| 51 | IMHO this is Sun's problem, not mine.
|
|---|
| 52 |
|
|---|
| 53 | Arnold Robbins
|
|---|
| 54 | arnold@gnu.org
|
|---|
| 55 |
|
|---|
| 56 | Solution (a) to Problem #2:
|
|---|
| 57 | ===========================
|
|---|
| 58 | Now fixed in the code via Paul Eggert's 2001-09-0 patch. See the
|
|---|
| 59 | ChangeLog.
|
|---|
| 60 |
|
|---|
| 61 | Solution (b) to Problem #2:
|
|---|
| 62 | ===========================
|
|---|
| 63 | From: Paul Nevai <nevai@math.ohio-state.edu>
|
|---|
| 64 | Subject: Re: gawk-3.0.4
|
|---|
| 65 | To: arnold@skeeve.com (Aharon Robbins)
|
|---|
| 66 | Date: Tue, 6 Jul 1999 09:09:05 -0400 (EDT)
|
|---|
| 67 |
|
|---|
| 68 | Dear Aharon:
|
|---|
| 69 |
|
|---|
| 70 | Toda raba. Why don't you add something like that to README_d/README.solaris
|
|---|
| 71 | file:
|
|---|
| 72 |
|
|---|
| 73 | for the SunOS do in
|
|---|
| 74 |
|
|---|
| 75 | /bin/sh: ulimit -n 256; ulimit -a; make test
|
|---|
| 76 | /bin/tcsh: limit descriptors 256; ulimit -a; make test
|
|---|
| 77 |
|
|---|
| 78 | otherwise "make test" will fail
|
|---|
| 79 |
|
|---|
| 80 | Shalom, Paul
|
|---|
| 81 |
|
|---|
| 82 | Aharon Robbins wrote to Paul Nevai:
|
|---|
| 83 | # >From the README_d/README.solaris file:
|
|---|
| 84 | #
|
|---|
| 85 | # Tue Apr 13 16:57:45 IDT 1999
|
|---|
| 86 | #
|
|---|
| 87 | # There is a known problem in that the `manyfiles' test will fail under
|
|---|
| 88 | # Solaris if you set your soft limit on the number of file descriptors to
|
|---|
| 89 | # above 256. This is due to a "feature" of fdopen that an fd must be
|
|---|
| 90 | # less than 256 (see fdopen(3)).
|
|---|
| 91 | #
|
|---|
| 92 | # IMHO this is Sun's problem, not mine.
|
|---|
| 93 | #
|
|---|
| 94 | # Arnold Robbins
|
|---|
| 95 | # arnold@gnu.org
|
|---|
| 96 | #
|
|---|
| 97 | # Double check your settings with ulimit; I suspect that this is
|
|---|
| 98 | # your problem.
|
|---|
| 99 | #
|
|---|
| 100 | # Thanks,
|
|---|
| 101 | #
|
|---|
| 102 | # Arnold
|
|---|
| 103 | # --
|
|---|
| 104 | # Aharon (Arnold) Robbins arnold@skeeve.com [ <<=== NOTE: NEW ADDRESS!! ]
|
|---|
| 105 | # P.O. Box 354 Home Phone: +972 8 979-0381 Fax: +1 603 761-6761
|
|---|
| 106 | # Nof Ayalon Cell Phone: +972 51 297-545 (See www.efax.com)
|
|---|
| 107 | # D.N. Shimshon 99784 Laundry increases exponentially in the
|
|---|
| 108 | # ISRAEL number of children. -- Miriam Robbins
|
|---|
| 109 | #
|
|---|
| 110 | #
|
|---|
| 111 |
|
|---|
| 112 |
|
|---|
| 113 |
|
|---|
| 114 | Paul Nevai pali+@osu.edu
|
|---|
| 115 | Department of Mathematics nevai@math.ohio-state.edu
|
|---|
| 116 | The Ohio State University http://www.math.ohio-state.edu/~nevai/
|
|---|
| 117 | 231 West Eighteenth Avenue http://www.math.ohio-state.edu/~jat/
|
|---|
| 118 | Columbus, Ohio 43210-1174 1-614-292-5310 (Office/Answering Device)
|
|---|
| 119 | The United States of America 1-614-292-1479 (Math Dept Fax)
|
|---|
| 120 |
|
|---|
| 121 | Solaris Problem #3:
|
|---|
| 122 | ===================
|
|---|
| 123 | Sun Feb 9 10:35:51 IST 2003
|
|---|
| 124 |
|
|---|
| 125 | Certain versions of Sun C give compilation errors under Solaris 5.5, 5.6 and
|
|---|
| 126 | possibly later. Here's what I was told:
|
|---|
| 127 |
|
|---|
| 128 | > We have this version of cc here:
|
|---|
| 129 | > cc -V
|
|---|
| 130 | > cc: Sun WorkShop 6 update 1 C 5.2 2000/09/11
|
|---|
| 131 | >
|
|---|
| 132 | > Probably, the others use different combinations of OS and CC.
|
|---|
| 133 | > A quick fix was this (we use csh-syntax here):
|
|---|
| 134 | >
|
|---|
| 135 | > setenv CC "/opt/SUNWspro/bin/cc -Xc"
|
|---|
| 136 | > ./configure
|
|---|
| 137 | > make check
|
|---|
| 138 |
|
|---|