Commit | Line | Data |
---|---|---|
920dae64 AT |
1 | If you read this file _as_is_, just ignore the funny characters you see. |
2 | It is written in the POD format (see pod/perlpod.pod) which is specially | |
3 | designed to be readable as is. | |
4 | ||
5 | =head1 NAME | |
6 | ||
7 | README.tru64 - Perl version 5 on Tru64 (formerly known as Digital UNIX formerly known as DEC OSF/1) systems | |
8 | ||
9 | =head1 DESCRIPTION | |
10 | ||
11 | This document describes various features of HP's (formerly Compaq's, | |
12 | formerly Digital's) Unix operating system (Tru64) that will affect | |
13 | how Perl version 5 (hereafter just Perl) is configured, compiled | |
14 | and/or runs. | |
15 | ||
16 | =head2 Compiling Perl 5 on Tru64 | |
17 | ||
18 | The recommended compiler to use in Tru64 is the native C compiler. | |
19 | The native compiler produces much faster code (the speed difference is | |
20 | noticeable: several dozen percentages) and also more correct code: if | |
21 | you are considering using the GNU C compiler you should use at the | |
22 | very least the release of 2.95.3 since all older gcc releases are | |
23 | known to produce broken code when compiling Perl. One manifestation | |
24 | of this brokenness is the lib/sdbm test dumping core; another is many | |
25 | of the op/regexp and op/pat, or ext/Storable tests dumping core | |
26 | (the exact pattern of failures depending on the GCC release and | |
27 | optimization flags). | |
28 | ||
29 | gcc 3.2.1 is known to work okay with Perl 5.8.0. However, when | |
30 | optimizing the toke.c gcc likes to have a lot of memory, 256 megabytes | |
31 | seems to be enough. The default setting of the process data section | |
32 | in Tru64 should be one gigabyte, but some sites/setups might have | |
33 | lowered that. The configuration process of Perl checks for too low | |
34 | process limits, and lowers the optimization for the toke.c if | |
35 | necessary, and also gives advice on how to raise the process limits. | |
36 | ||
37 | =head2 Using Large Files with Perl on Tru64 | |
38 | ||
39 | In Tru64 Perl is automatically able to use large files, that is, | |
40 | files larger than 2 gigabytes, there is no need to use the Configure | |
41 | -Duselargefiles option as described in INSTALL (though using the option | |
42 | is harmless). | |
43 | ||
44 | =head2 Threaded Perl on Tru64 | |
45 | ||
46 | If you want to use threads, you should primarily use the new Perl | |
47 | 5.8.0 threads model by running Configure with -Duseithreads. | |
48 | ||
49 | The old Perl 5.005 threads is obsolete, unmaintained, and its use is | |
50 | discouraged. If you really want it, run Configure with the | |
51 | -Dusethreads -Duse5005threads options as described in INSTALL. | |
52 | ||
53 | Either thread model is going to work only in Tru64 4.0 and newer | |
54 | releases, older operating releases like 3.2 aren't probably going | |
55 | to work properly with threads. | |
56 | ||
57 | In Tru64 V5 (at least V5.1A, V5.1B) you cannot build threaded Perl with gcc | |
58 | because the system header <pthread.h> explicitly checks for supported | |
59 | C compilers, gcc (at least 3.2.2) not being one of them. But the | |
60 | system C compiler should work just fine. | |
61 | ||
62 | =head2 Long Doubles on Tru64 | |
63 | ||
64 | You cannot Configure Perl to use long doubles unless you have at least | |
65 | Tru64 V5.0, the long double support simply wasn't functional enough | |
66 | before that. Perl's Configure will override attempts to use the long | |
67 | doubles (you can notice this by Configure finding out that the modfl() | |
68 | function does not work as it should). | |
69 | ||
70 | At the time of this writing (June 2002), there is a known bug in the | |
71 | Tru64 libc printing of long doubles when not using "e" notation. | |
72 | The values are correct and usable, but you only get a limited number | |
73 | of digits displayed unless you force the issue by using C<printf | |
74 | "%.33e",$num> or the like. For Tru64 versions V5.0A through V5.1A, a | |
75 | patch is expected sometime after perl 5.8.0 is released. If your libc | |
76 | has not yet been patched, you'll get a warning from Configure when | |
77 | selecting long doubles. | |
78 | ||
79 | =head2 DB_File tests failing on Tru64 | |
80 | ||
81 | The DB_File tests (db-btree.t, db-hash.t, db-recno.t) may fail you | |
82 | have installed a newer version of Berkeley DB into the system and the | |
83 | -I and -L compiler and linker flags introduce version conflicts with | |
84 | the DB 1.85 headers and libraries that came with the Tru64. For example, | |
85 | mixing a DB v2 library with the DB v1 headers is a bad idea. Watch | |
86 | out for Configure options -Dlocincpth and -Dloclibpth, and check your | |
87 | /usr/local/include and /usr/local/lib since they are included by default. | |
88 | ||
89 | The second option is to explicitly instruct Configure to detect the | |
90 | newer Berkeley DB installation, by supplying the right directories with | |
91 | C<-Dlocincpth=/some/include> and C<-Dloclibpth=/some/lib> B<and> before | |
92 | running "make test" setting your LD_LIBRARY_PATH to F</some/lib>. | |
93 | ||
94 | The third option is to work around the problem by disabling the | |
95 | DB_File completely when build Perl by specifying -Ui_db to Configure, | |
96 | and then using the BerkeleyDB module from CPAN instead of DB_File. | |
97 | The BerkeleyDB works with Berkeley DB versions 2.* or greater. | |
98 | ||
99 | The Berkeley DB 4.1.25 has been tested with Tru64 V5.1A and found | |
100 | to work. The latest Berkeley DB can be found from F<http://www.sleepycat.com>. | |
101 | ||
102 | =head2 64-bit Perl on Tru64 | |
103 | ||
104 | In Tru64 Perl's integers are automatically 64-bit wide, there is | |
105 | no need to use the Configure -Duse64bitint option as described | |
106 | in INSTALL. Similarly, there is no need for -Duse64bitall | |
107 | since pointers are automatically 64-bit wide. | |
108 | ||
109 | =head2 Warnings about floating-point overflow when compiling Perl on Tru64 | |
110 | ||
111 | When compiling Perl in Tru64 you may (depending on the compiler | |
112 | release) see two warnings like this | |
113 | ||
114 | cc: Warning: numeric.c, line 104: In this statement, floating-point overflow occurs in evaluating the expression "1.8e308". (floatoverfl) | |
115 | return HUGE_VAL; | |
116 | -----------^ | |
117 | ||
118 | and when compiling the POSIX extension | |
119 | ||
120 | cc: Warning: const-c.inc, line 2007: In this statement, floating-point overflow occurs in evaluating the expression "1.8e308". (floatoverfl) | |
121 | return HUGE_VAL; | |
122 | -------------------^ | |
123 | ||
124 | The exact line numbers may vary between Perl releases. The warnings | |
125 | are benign and can be ignored: in later C compiler releases the warnings | |
126 | should be gone. | |
127 | ||
128 | When the file F<pp_sys.c> is being compiled you may (depending on the | |
129 | operating system release) see an additional compiler flag being used: | |
130 | C<-DNO_EFF_ONLY_OK>. This is normal and refers to a feature that is | |
131 | relevant only if you use the C<filetest> pragma. In older releases of | |
132 | the operating system the feature was broken and the NO_EFF_ONLY_OK | |
133 | instructs Perl not to use the feature. | |
134 | ||
135 | =head1 Testing Perl on Tru64 | |
136 | ||
137 | During "make test" the C<comp/cpp> will be skipped because on Tru64 it | |
138 | cannot be tested before Perl has been installed. The test refers to | |
139 | the use of the C<-P> option of Perl. | |
140 | ||
141 | =head1 ext/ODBM_File/odbm Test Failing With Static Builds | |
142 | ||
143 | The ext/ODBM_File/odbm is known to fail with static builds | |
144 | (Configure -Uusedl) due to a known bug in Tru64's static libdbm | |
145 | library. The good news is that you very probably don't need to ever | |
146 | use the ODBM_File extension since more advanced NDBM_File works fine, | |
147 | not to mention the even more advanced DB_File. | |
148 | ||
149 | =head1 Perl Fails Because Of Unresolved Symbol sockatmark | |
150 | ||
151 | If you get an error like | |
152 | ||
153 | Can't load '.../OSF1/lib/perl5/5.8.0/alpha-dec_osf/auto/IO/IO.so' for module IO: Unresolved symbol in .../lib/perl5/5.8.0/alpha-dec_osf/auto/IO/IO.so: sockatmark at .../lib/perl5/5.8.0/alpha-dec_osf/XSLoader.pm line 75. | |
154 | ||
155 | you need to either recompile your Perl in Tru64 4.0D or upgrade your | |
156 | Tru64 4.0D to at least 4.0F: the sockatmark() system call was | |
157 | added in Tru64 4.0F, and the IO extension refers that symbol. | |
158 | ||
159 | =head1 AUTHOR | |
160 | ||
161 | Jarkko Hietaniemi <jhi@iki.fi> | |
162 | ||
163 | =cut |