[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
potential AC_LIBTOOL_SYS_MAX_CMD_LEN bug
From: |
Willem Jan Palenstijn |
Subject: |
potential AC_LIBTOOL_SYS_MAX_CMD_LEN bug |
Date: |
Wed, 4 Aug 2004 00:01:22 +0200 (CEST) |
Hi,
I encountered a problem today while running a configure (from GemRB
(gemrb.sf.net)) generated by autoconf 2.59 with a libtool.m4 from libtool
1.5.2.
I started configure with "bash configure" and did not have
$CONFIG_SHELL set. Note that in this case $0 equals "configure".
configure contains the following test from libtool.m4 in the
AC_LIBTOOL_SYS_MAX_CMD_LEN macro:
test "X"`$CONFIG_SHELL $0 --fallback-echo "X$testring" 2>/dev/null` ==
"XX$testring"
This test will always fail, since "$CONFIG_SHELL $0" expands to
" configure", which is not in my path (since "." is not in my $PATH). This
results in max_cmd_len being empty, which is clearly undesirable.
I double-checked this by adding the line:
echo `$CONFIG_SHELL $0 --fallback-echo "XX" 2>/dev/null`
to configure just above the test, and it indeed produced an empty line
when configure was run as "bash configure" and "XX", as expected,
when run as "./configure".
In other words, I do not think $CONFIG_SHELL can be used for this
purpose at this point in libtool.m4.
If you need any other info, I'll be happy to provide it.
Thanks,
Willem Jan
- potential AC_LIBTOOL_SYS_MAX_CMD_LEN bug,
Willem Jan Palenstijn <=