| Akim,
| stenn@pogo sh
| $ foo=`cd / 2/dev/null`
| $ echo ,$bar,
| ,,
| $ baz=`exit nonsense 2/dev/null`
| $ echo ,$baz,
| ,,
| $
Thanks, but you skipped the most important tests :)
Could you try all of these?
foo=`cd / 2/dev/null`
echo ,$foo,
bar=`cd /foo 2/dev/null`
echo ,$bar,
baz=`exit
On Tue, Jan 30, 2001 at 09:09:53PM -0500, Harlan Stenn wrote:
: Akim,
:
: stenn@pogo sh
: $ foo=`cd / 2/dev/null`
: $ echo ,$bar,
: ,,
In case noone else sees it, you assigned to $foo and output $bar, so
the result is naturally empty. You should probably try again without
mixing test 1 and 2
Sigh.
ULTRIX V4.4 (Rev. 69) System #31: Thu Aug 10 19:42:23 GMT 1995
UWS V4.4 (Rev. 11)
No mail.
stenn@porkypine sh
$ foo=`cd / 2/dev/null`
illegal io
$ echo ,$foo,
,,
$ bar=`cd /foo 2/dev/null`
/foo: bad directory
$ echo ,$bar,
,,
$ baz=`exit nonsense 2/dev/null`
nonsense: bad number
$ echo
"Harlan" == Harlan Stenn [EMAIL PROTECTED] writes:
Harlan The tail fragment of "make check" follows. Much better - I'll
Harlan see if I can install the GNU grep stuff and see if that helps.
When make check fails building the testsuite, just run it again, it
should work the second time. I
| $ at_dir=`echo foo 2/dev/null`
| $ echo at_dir=$at_dir
| at_dir=foo
I don't understand this, it is not consistent with `cd'. Isn't `echo'
a builtin on your shell? Then I would understand.
Hm, what else must be a builtin? exit has to be.
foo=`exit 12 2/dev/null`
bar=`(exit 12) 2/dev/null`
Akim,
I just downloaded a fresh CVS autoconf and did a build on the Ultrix
box.
This time, I was able to do the build in a subdir (builddir != srcdir).
The tail fragment of "make check" follows. Much better - I'll see if I
can install the GNU grep stuff and see if that helps.
Harlan
-
...
"Harlan" == Harlan Stenn [EMAIL PROTECTED] writes:
Harlan Akim, Here's the sh -x testsuite snippet.
Thanks, this is the culprit:
at_dir=`cd "$at_dir" 2/dev/null pwd`
because PATH may include nonexistent dirs, I included the 2/dev/null
part.
Can you try these please? I'd like to know if
Akim,
Here they are.
Harlan
---
Can you try these please? I'd like to know if the problem is with the
assign or the substitution, etc.
$ at_dir=`(cd /foo pwd) 2/dev/null`
$ echo at_dir=$at_dir
at_dir=
$ at_dir=`(cd /foo pwd)` 2/dev/null
/foo: bad directory
$ echo at_dir=$at_dir
at_dir=
$
| Akim,
| I installed that version of autoconf and put your script in a file
| called "foo":
|
| stenn@porkypine sh foo
| configure: loading cache /dev/null
| configure: creating ./config.status
| config.status: creating file
| config.status: creating header.h
| BAR
| /* header.h. Generated
Akim,
stenn@porkypine sh foo
configure: loading cache /dev/null
configure: creating ./config.status
config.status: creating file1
config.status: creating file2
BAR
BAR
stenn@porkypine
Harlan
| Akim,
Hi! Thanks Harlan!
| chmod +x testsuite-tmp
| mv testsuite-tmp testsuite
| /bin/sh testsuite
| /bin/sh: illegal io
| /bin/sh: illegal io
| /bin/sh: illegal io
That was to be expected. If you could send a snippet of sh -x
testsuite maybe we could fix the test suite on Ultrix, which
Akim,
I installed that version of autoconf and put your script in a file
called "foo":
stenn@porkypine sh foo
configure: loading cache /dev/null
configure: creating ./config.status
config.status: creating file
config.status: creating header.h
BAR
/* header.h. Generated automatically by
Akim,
Here's the sh -x testsuite snippet.
Harlan
---
...
+ cd ..
+ pwd
at_top_srcdir=/home/stenn/src/gnu/autoconf
+ pwd
at_path=/home/stenn/src/gnu/autoconf/tests
at_IFS_save=
IFS=:
+ cd ..
sh: illegal io
at_dir=
+ test -n
+ cd /usa/stenn/bin
sh: illegal io
at_dir=
+ test -n
+ cd
Akim,
I grabbed the CVS copy a few minutes ago - it looked like your patch had
been applied.
I did a "configure/make/make check" on the Ultrix box.
I originally tried to do the build/test in a subdir, but that failed
trying to locate some of the files for sed processing (as I recall).
Harlan
"Kevin" == Kevin Ryde [EMAIL PROTECTED] writes:
Kevin Alas attempts at making such an input have met with miserable
Kevin failure.
Kevin Whatever the bug is, it seems highly dependent on the
Kevin environment. I got to config.status.cutdown below which
Kevin provokes the segv under
Kevin
"akim" == akim [EMAIL PROTECTED] writes:
akim Thanks. We must be hitting some internal size limitation I
akim guess. I will apply the patch you suggest.
I'm applying the patch below. Harlan, please, try this Autoconf, or
maybe even the test suite (but I'm quite pessimistic wrt the
Thanks. We must be hitting some internal size limitation I guess. I will
apply the patch you suggest.
Akim Demaille [EMAIL PROTECTED] writes:
"Torbjorn" == Torbjorn Granlund [EMAIL PROTECTED] writes:
Torbjorn Running configure from the current GMP working sources,
Torbjorn SunOS 4.1.3 /bin/sh catches a segfault.
Thanks a lot. I'm working on it, which means I first try to get this
*(*$
"Torbjorn" == Torbjorn Granlund [EMAIL PROTECTED] writes:
Torbjorn Running configure from the current GMP working sources,
Torbjorn SunOS 4.1.3 /bin/sh catches a segfault.
Thanks a lot. I'm working on it, which means I first try to get this
*(*£$ test suite work properly on SunOS. I don't
Running configure from the current GMP working sources,
SunOS 4.1.3 /bin/sh catches a segfault.
I tracked down the problem to some lines in the newly generated
config.status:
# If the user did not use the arguments to specify the items to instantiate,
# then the envvar interface is used.
20 matches
Mail list logo