Don't know if this is a related problem (using make -e with a different
SHEL var does not make any diff so I guess not). Anyway I get the
following a couple of times during 'gmake install':
...
dumped_data found at offset 0x346fb8, patching.
dumped_data found at offset 0x346fb8, patching.
if ! test -f dump-size; then \
../lib-src/insert-data-in-exec -s xemacs.dmp > dump-size ; \
rm -f dump-data.o xemacs xemacs.dmp temacs;\
gmake xemacs; \
fi
/bin/sh: !: execute permission denied
Installation works but it makes me a bit nervous.
/MOL
On 2005-02-25 09:17, Sylvain Mazet wrote:
make -e
does it.
Thanks.
Sylvain.
>>>>>"Vin" == Vin Shelton <ethersoft(a)gmail.com> writes:
Vin> On Thu, 24 Feb 2005 11:36:34 +0100, Sylvain Mazet
Vin> <sylvain.mazet(a)silvaco.com> wrote:
>> Hi,
>>
>> compiling XEmacs-21.5.19 on Sun (solaris 5.8), I get this error
>> when gnumaking:
>>
>> --------------------------------------------------
>> Dumping under the name xemacs if test -f dump-size; then \
>> ../lib-src/insert-data-in-exec temacs xemacs.dmp xemacs `
>> ./temacs -si`; \ ret=$? ; \ if test ${ret} -eq 2; then \ rm -f
>> dump-size ; \ else \ if test ${ret} -eq 1; then \ exit 1; \
>> else \ chmod +x xemacs ; \ fi ; \ fi ; \ fi dumped_data found
>> at offset 0x45c208, patching. dumped_data found at offset
>> 0x45c208, patching. if ! test -f dump-size; then \
>> ../lib-src/insert-data-in-exec -s xemacs.dmp > dump-size ; \ rm
>> -f dump-data.o xemacs xemacs.dmp temacs;\ gnumake xemacs; \ fi
>> /bin/sh: !: not found
>> --------------------------------------------------
>>
>> And then compilation goes on to completion....
>>
>> I did set CONFIG_SHELL=/usr/xpg4/bin/sh in my environment
>> before configuring.
>>
>> I found that changing SHELL=/bin/sh to SHELL=/usr/xpg4/bin/sh
>>
>> in src/GNUmakefile removes the error.
>>
>> It seems to me that SHELL=/bin/sh is hard coded in the
>> Makefile.in.in files, and triggers the same kind of problem as
>> for the configure script.
>>
>> I can provide more info if needed,
>>
>> Sylvain Mazet.
>>
>>
Vin> Does running 'make -e' help? That should force the
Vin> environment setting of SHELL to take priority.