zsh-4.1.1-test-1 on HP-UX 11

zsh-4.1.1-test-1 on HP-UX 11

[Paul Ackersviller]

The test release has been working well for me on HP-UX, but I'm
getting a few failures on the tests because I have no zsh/zpty.
However I never saw any configuration or compilation errors due to
this -- is zpty supposed to get left out on HP-UX?

Regarding the configuration, it seems --disable-dynamic is necessary
on this system, at least with Softbench (I haven't tried gcc yet).
Also the remark about -lHcurses in Etc/MACHINES doesn't seem needed
anymore, since this is done correctly now.

-- 
Paul Ackersviller

Re: zsh-4.1.1-test-1 on HP-UX 11

[Bart Schaefer]

On May 1,  5:47pm, Paul Ackersviller wrote:
} Subject: zsh-4.1.1-test-1 on HP-UX 11
}
} The test release has been working well for me on HP-UX, but I'm
} getting a few failures on the tests because I have no zsh/zpty.
[...]
} Regarding the configuration, it seems --disable-dynamic is necessary

The reason you don't have zpty is because it's only loaded dynamically,
unless you edit the modules.conf file.

Re: zsh-4.1.1-test-1 on HP-UX 11

[Peter Stephenson]

Paul Ackersviller wrote:
> The test release has been working well for me on HP-UX, but I'm
> getting a few failures on the tests because I have no zsh/zpty.
> However I never saw any configuration or compilation errors due to
> this -- is zpty supposed to get left out on HP-UX?

> Regarding the configuration, it seems --disable-dynamic is necessary
> on this system, at least with Softbench (I haven't tried gcc yet).

I'll put this in the MACHINES file.  I think Bart's point is that these
two issues are tied --- if you --disable-dynamic you need to change
`link=dynamic' to `link=static' on the zpty line in config.modules for
that to work.  However, zpty is particularly configuration dependent so
it's not impossible it needs tweaking.

Index: Etc/MACHINES
===================================================================
RCS file: /cvsroot/zsh/zsh/Etc/MACHINES,v
retrieving revision 1.17
diff -u -r1.17 MACHINES
--- Etc/MACHINES	14 Feb 2003 15:57:03 -0000	1.17
+++ Etc/MACHINES	2 May 2003 09:33:02 -0000
@@ -79,11 +79,15 @@
 	  is reported to fix.  An attempt to fix this in configure
 	  is apparently ineffective; more information would be appreciated
 	  as the maintainers do not have access to an HP-UX system.
+	  Recent reports indicated this is not necessary on recent versions
+	  of HP-UX 11.
 
-	  Problems with dynamic loading have been reported, but
-	  this should compile using the standard dlopen() function set
-	  (rather than the 10.20 shl_load() function set).  More details of
-	  any difficulties would be appreciated.
+	  The configure option `--disable-dynamic' may be necessary when
+	  compiling on HP-UX 11 using Softbench.  Various problems with
+	  dynamic loading have been reported; this should compile using the
+	  standard dlopen() function set (rather than the 10.20 shl_load()
+	  function set).  More details of any difficulties would be
+	  appreciated.
 
 	  Compiling with gcc 2.7.1 is known to fail with header file
 	  conflicts.  Use the HP ANSI C compiler.

-- 
Peter Stephenson <pws@csr.com>                  Software Engineer
CSR Ltd., Science Park, Milton Road,
Cambridge, CB4 0WH, UK                          Tel: +44 (0)1223 692070


**********************************************************************
The information transmitted is intended only for the person or
entity to which it is addressed and may contain confidential 
and/or privileged material. 
Any review, retransmission, dissemination or other use of, or
taking of any action in reliance upon, this information by 
persons or entities other than the intended recipient is 
prohibited.  
If you received this in error, please contact the sender and 
delete the material from any computer.
**********************************************************************

Re: zsh-4.1.1-test-1 on HP-UX 11

[Paul Ackersviller]

On Fri, May 02, 2003 at 02:34:58AM +0000, Bart Schaefer wrote:
> On May 1,  5:47pm, Paul Ackersviller wrote:
> } Regarding the configuration, it seems --disable-dynamic is necessary
> 
> The reason you don't have zpty is because it's only loaded dynamically,
> unless you edit the modules.conf file.

Thanks, that does get it made for me, however Y01completion.ztst appears
to hang indefinitely on testing.  ZTST_verbose=2 make test gives me:
    ./Y01completion.ztst: starting.
    ZTST_getsect: read section name: prep
    ZTST_getchunk: read code chunk:
    . $ZTST_srcdir/comptest
    ZTST_execchunk: status 0
    ZTST_getchunk: read code chunk:
    mkdir comp.tmp
    cd comp.tmp
    ZTST_execchunk: status 0
    ZTST_getchunk: read code chunk:
    comptestinit -z $ZTST_testdir/../Src/zsh &&
    {
        mkdir dir1 &&
        mkdir dir2 &&
        touch file1 &&
        touch file2
    }


I assume then that HP-UX is one of the known systems on which dynamic
loading doesn't work, true?  Has there been any discussion on this
in the past that I could refer to, and does anyone recall whether the
the problems look fixable?

Regarding Peter's followup, I had a chance to try gcc 3.0.2 on zsh and
had no trouble -- it gives me identical results to HP's compiler for
everything I've tried.

-- 
Paul Ackersviller

Re: zsh-4.1.1-test-1 on HP-UX 11

[Paul Ackersviller]

On Mon, May 05, 2003 at 12:38:45PM -0400, Paul Ackersviller wrote:
> I assume then that HP-UX is one of the known systems on which dynamic
> loading doesn't work, true?  Has there been any discussion on this
> in the past that I could refer to, and does anyone recall whether the
> the problems look fixable?

I expect I'll be able to answer my own question now that I see mailing
list searches are working.

Am I missing something, or is there not a way to do searches in
reverse chronological order?  This would be more user-friendly.

-- 
Paul Ackersviller