Commit Graph

1086 Commits

Author SHA1 Message Date
tg
180883417e tentatively commit support for
• detecting a lot more compilers (stolen from CMake, admittedly, as they
  were the first coming up at G**gle with usable #ifdefs)
• handling MIPSpro
• more system info on IRIX
2008-03-27 22:17:01 +00:00
tg
351104715e ignore failure to define MAP_FILE 2008-03-27 17:59:28 +00:00
tg
dbc159a491 From: Elias Pipping <elias@pipping.org>
| >Maybe add <sys/types.h> as first line? I could imagine that.
|
| That did the trick indeed (that's where dev_t is defined).
2008-03-27 17:55:31 +00:00
tg
5ff5b4b03f Elias Pipping reports success of mksh R33 on IRIX64 mips-sgi-irix6.5 gcc3.4.1 2008-03-27 13:08:37 +00:00
tg
def9c172df • Add support for Ultrix 4.5 and ucode cc (?)
‣ I/O redirection seems broken:
    $ (date; date >/dev/null; date) | wc -l
    1 (expected: 2)
  ‣ other than that: working fine
  ‣ -YBSD (default) and -YSYSTEM_FIVE don’t work, just -YPOSIX, somehow
• Fix $(…) to `…` for OSF/1 V2.0 /bin/sh
  ‣ this compiler is FUBAR though:
	$ cat >t.c
	main() { return (foo()); }
	$ cc t.c
	ld:
	Unresolved :
	foo
	$ echo $?
	0
	$ ls -l a.out
	-rwxr-xr-x   1 mirbsd   users      10835 Jul 21 17:12 a.out
  ‣ it seems to have ucode, but man is not installed
• new mirtoconf check: mkstemp(3)
• if !HAVE_MKSTEMP (Ultrix), use tempnam(3)
• only use printf(1) if it exists (it doesn’t on Ultrix)
• a few more signals
• add S_ISLNK if the OS doesn’t define it
• add strcasecmp(3) proto for Ultrix (it _is_ in <portability.h>, but
  only for -YBSD I think)
• fgrep(1) on Ultrix doesn’t do “-e ① -e ②”

10x DEChengst:#UnixNL for giving access
2008-03-25 21:34:45 +00:00
tg
b8b52a7c92 set MI default values after MD default values 2008-03-25 20:25:27 +00:00
tg
83b8798da3 support dæmonisation in mksh, for example
|	csh -cf '/command/svscanboot &'
and
|	/usr/mpkg/bin/pgrphack /usr/mpkg/bin/svscanboot &
can now be replaced with
|	/bin/mksh -T- /usr/mpkg/bin/svscanboot
2008-03-23 22:09:59 +00:00
tg
38a53603d3 support gcc 1.42: it only can __attribute__((aligned (1))) but no others,
it can’t do -Werror, that’s why it mis-detects attribute support.
ignore the warnings (unused variables, uninitialised values, control flow)
as I checked them and they are not valid. gcc 1.42 (BSDi modified) compiles
a very fine mksh
2008-03-23 21:53:36 +00:00
tg
425410de19 make rcs ids ident(1) capable 2008-03-23 21:31:29 +00:00
tg
26ba2854a2 support BSD/OS 3.1 with gcc2, 10x replaced@tnf
except MAP_FAILED not being defined, no issues
2008-03-23 20:55:18 +00:00
tg
d18e58e74d fix environment handling for perl 5.003_02
gotta love 1997’s software
2008-03-23 20:54:29 +00:00
tg
bfe7bff954 add regression test self-tests 2008-03-23 20:43:51 +00:00
tg
df9b1b9b84 choose perl5 over perl, which can be perl 4.0… 2008-03-23 20:21:24 +00:00
tg
444214f5c2 fix Rebuild.sh generation 2008-03-23 20:19:26 +00:00
tg
5f22a5cd7c much improve the Syllable situation – now en par with Plan 9 2008-03-14 21:34:55 +00:00
tg
17ffadce08 attempt version report on msc/interix 2008-03-06 18:12:58 +00:00
tg
a9501a7372 more hw info, 10x penpen 2008-03-06 18:02:33 +00:00
tg
34aa8c7a8b make gcc not nōn-silently accept -std=gnu99 if it can’t 2008-03-05 19:59:27 +00:00
tg
f0234c03b9 -Wl,+k is no longer needed on HP-sUX with the native linker now that we
check the errorlevel
XXX TODO: Digital Mars
2008-03-05 19:14:19 +00:00
tg
845f6f0d64 • check for flock decl too (weird on OSF/1: if !POSIX and BSD)
• un-experimental Tru64
2008-03-05 18:49:15 +00:00
tg
194972394d yeah, sure… 2008-03-05 18:25:57 +00:00
tg
0e6df4736a • now this builds fine on DEChengst’s Tru64 box:
| OSF1 rubbereendje.dechengst.nl V5.1 2650 alpha
  with the vendor compiler:
  | Compaq C V6.5-011 on HP Tru64 UNIX V5.1B (Rev. 2650)
  | Compiler Driver V6.5-003 (sys) cc Driver
• the platform’s sig_t is incompatible too (simplify check)
• no compile warnings at all
• results in:
  $ size mksh
  | text    data    bss     dec     hex
  | 327680  16384   17808   361872  58590
  $ file mksh
  | mksh:   COFF format alpha dynamically linked, demand paged executable or object module not stripped - version 3.13-14
  $ ldd mksh
  |
  |         Main  =>   mksh
  |         libc.so  =>   /usr/shlib/libc.so
  $ ls -l mksh
  | -rwxr-xr-x   1 mirbsd   users     395200 Mar  5 19:18 mksh
• minor testsuite issues:
  FAIL ./check.t:regression-13
        unexpected stderr - got too much output
        wanted nothing
        got:
                Successful
                cat: output error
  ⇒ probably harmless
• works like a charm!
2008-03-05 18:21:45 +00:00
tg
a592c65cf0 somehow magically make it seem to work 2008-03-05 17:52:43 +00:00
tg
7081a38796 Tru64 aka OSF/1 5.1 (10x DEChengst@#UnixNL) is more nice than OSF/1 4.0 2008-03-05 17:30:10 +00:00
tg
f25151387d • OSF/1, PW32: speed up mirtoconf test for sig_t (which does not exist)
• OSF/1: need _XOPEN_SOURCE_EXTENDED for mknod(2) decl
2008-03-05 17:12:08 +00:00
tg
3d9bd0e1f0 OSF/1 doesn’t seem to declare revoke(2) anywhere 2008-03-05 17:06:50 +00:00
tg
227a5af8bc try to support
• DEC C on OSF/1 (10x Jupp the IceWM coffee pot maintainer)
• other stuff which doesn’t nuke a.out on failure

XXX this must be tested on *ALL* supported platforms!
2008-03-05 17:00:33 +00:00
tg
94158b2326 shut up gcc-lib/alpha-dec-osf4.0d/2.8.1 2008-03-05 16:54:21 +00:00
tg
3321382ad2 more, this time on mirbsd… *sigh* 2008-03-05 16:49:22 +00:00
tg
3ff6cc397a found by gcc on OSF/1 10x to Jupp Schugt (penpen) 2008-03-05 16:40:57 +00:00
tg
231b682cdf according to groff_char(7) and my own hard experiences, there is more wrong
with GNU groff – add some special handling to the BSD mdoc macros for it so
that the manual pages look good in both utf8 and ps (PDF) mode; also fix in
mksh wrong display of ` (groff: ‘), ' (groff: ’), \' (groff: ´), \- (groff:
U+2212 −), the en dash (nroff doesn’t have it, use the em dash there ONLY),
and ~ and ^ (groff: placed atop and size-reduced, for use as diacritics, in
manual pages bad since these are control characters there)
→ PDF manpage now has ‘’ “” and good-looking hyphens and mini and ~ and ^
→ utf8 manpage now has ‘’ “”, good-looking hyphens, cut’n’pasteable mini
→ nroff manpage still has '' "" instead of ugly `' ``'' or even `´
→ Debian lintian won’t complain any longer
2008-03-04 00:09:34 +00:00
tg
8a3169a0dd Patrick “aptituz” Schoenfeld mentioned “lintian -viI mksh*.changes” to
me, which points out that “gnroff -Tutf8” mangles the ‘-’ characters
(hyphen/minus) from the input into ‘‐’ characters (hyphen), which does
not make sense in many cases and prevent copy’n’paste → fix

no change in nrcon output
2008-03-03 22:38:41 +00:00
tg
09318a472b no longer make the DIRSTACKBASE variable read-only – it makes too much
trouble when re-reading .mkshrc, and “we sell rope” after all
2008-03-03 20:22:37 +00:00
tg
2cae6c9ab5 report OE and compiler versions inside the build script now, where methods
to query them are known (missing: bcc/dmc/msc compilers; additional OS info
besides “uname -a” and Darwin’s “hwprefs os_type” – please tell me!)
2008-03-03 19:40:08 +00:00
tg
047da7a0eb simplify 2008-03-03 18:44:32 +00:00
tg
84debe1e7d add a much more crazy test now, for several occurences of escaped and
unescaped and variabled patterns and slashes
2008-03-01 22:58:22 +00:00
tg
dbae22571e a better way to check for empty patterns
(not quite how bash handles it - they do a count limit)
2008-03-01 22:49:37 +00:00
tg
947bc8e05b finally, all bugs fixed, adjust dot.mkshrc to use the subst code 2008-03-01 21:24:58 +00:00
tg
4912b5cf67 quite a big change, but now the variables expanded are not
scanned for slashes, plus the ADELIM code gets more use and
a bugfix ☺
2008-03-01 21:10:26 +00:00
tg
16dd1d3c8f fix one of the bugs in this code… still the one I’m looking for left 2008-03-01 17:14:17 +00:00
tg
707a787edc add another two corner case checks - and mksh fails one of these oO 2008-03-01 16:40:57 +00:00
tg
015e334ccf oops. forgot to commit:
• bump © year
• document better too
• optimise
2008-03-01 16:35:30 +00:00
tg
fe612d223b better debugging aid: print versions at several places 2008-03-01 15:07:50 +00:00
tg
90264a8b14 fix compilation in corner case, my bad 2008-03-01 13:57:36 +00:00
tg
36c4552e1f we don't need to be special if called as -sh any longer now 2008-03-01 02:21:38 +00:00
tg
f87b33997a two small fixes: ① cvs doesn’t like me; ② .Nm only takes one word argument 2008-02-29 18:20:16 +00:00
tg
8d170a3d01 fix on Interix, where tr(1) is more weird than even Solaris’ XPG4 one…
just do not use ranges, no matter what.
2008-02-29 16:38:41 +00:00
tg
c7594709db mention only one -e works 2008-02-29 12:48:09 +00:00
tg
20f32015d0 icc needs this 2008-02-29 12:47:03 +00:00
tg
d17e25c7f4 fix passing of env var to regression tests 2008-02-29 11:57:30 +00:00