vim 7.3

classic Classic list List threaded Threaded
10 messages Options
Reply | Threaded
Open this post in threaded view
|

vim 7.3

Charles Campbell
Hello,

I just attempted to update my copy of vim 7.2 with:

       (date && hg pull -u) 2>&1 |tee -a ../hg-vim.log

I followed that up with a   gmake distclean
and                                     ./configure --with-features=huge
--enable-gui=gtk2 --enable-perlinterp --enable-pythoninterp
--enable-gnome-check --enable-cscope

Configure failed with the following message:

checking size of off_t... configure: error: in
`/home/cec/.SW/VIM/Merc/vim/src':
configure: error: cannot compute sizeof (off_t)

Here's some excerpts from config.log:  (a centos 5.2 system)

hostname = xorn.gsfc.nasa.gov
uname -m = x86_64
uname -r = 2.6.18-194.3.1.el5
uname -s = Linux
uname -v = #1 SMP Thu May 13 13:08:30 EDT 2010

/usr/bin/uname -p = unknown
/bin/uname -X     = unknown

/bin/arch              = x86_64
/usr/bin/arch -k       = unknown
/usr/convex/getsysinfo = unknown
/usr/bin/hostinfo      = unknown
/bin/machine           = unknown
/usr/bin/oslevel       = unknown
/bin/universe          = unknown

...

configure:11492: checking size of off_t
configure:11497: gcc -o conftest -g -O2   -Wl,-E
-Wl,-rpath,/usr/lib64/perl5/5.8.8/x86_64-linux-thread-multi/CORE  
-L/usr/local/lib conftest.c -lm -lncurses -lelf -lnsl  -lselinux  -lacl
-lattr -lgpm >&5
In file included from /usr/include/inttypes.h:28,
                  from conftest.c:164:
/usr/include/stdint.h:52: error: duplicate 'unsigned'
/usr/include/stdint.h:52: error: two or more data types in declaration
specifiers
configure:11497: $? = 1
configure: program exited with status 1

Here's an excerpt from /usr/include/stdint.h, starting at line 51:

#ifndef __uint32_t_defined
typedef unsigned int        uint32_t;
# define __uint32_t_defined
#endif

Vim 7.3 configured&compiled ok before this pull, and FWIW vim 7.2
configures and builds ok (with patches 1-444).

Regards,
Chip Campbell

--
You received this message from the "vim_dev" maillist.
Do not top-post! Type your reply below the text you are replying to.
For more information, visit http://www.vim.org/maillist.php
Reply | Threaded
Open this post in threaded view
|

Re: vim 7.3

Charles Campbell
Charles Campbell wrote:
> Hello,
>
> I just attempted to update my copy of vim 7.2 with:
>
>       (date && hg pull -u) 2>&1 |tee -a ../hg-vim.log
>
sorry, that should've been  "...my copy of vim 7.3 with:..."

Regards,
Chip Campbell


--
You received this message from the "vim_dev" maillist.
Do not top-post! Type your reply below the text you are replying to.
For more information, visit http://www.vim.org/maillist.php
Reply | Threaded
Open this post in threaded view
|

Re: vim 7.3

Charles Campbell
In reply to this post by Charles Campbell
Charles Campbell wrote:

> Hello,
>
> I just attempted to update my copy of vim 7.2 with:
>
>       (date && hg pull -u) 2>&1 |tee -a ../hg-vim.log
>
> I followed that up with a   gmake distclean
> and                                     ./configure
> --with-features=huge --enable-gui=gtk2 --enable-perlinterp
> --enable-pythoninterp --enable-gnome-check --enable-cscope
>
> Configure failed with the following message:
>
and  it also fails the same way with a plain   ./configure

Regards,
Chip Campbell


--
You received this message from the "vim_dev" maillist.
Do not top-post! Type your reply below the text you are replying to.
For more information, visit http://www.vim.org/maillist.php
Reply | Threaded
Open this post in threaded view
|

Re: vim 7.3

James Vega-3
In reply to this post by Charles Campbell
On Tue, Jun 22, 2010 at 3:03 PM, Charles Campbell
<[hidden email]> wrote:

> Configure failed with the following message:
>
> checking size of off_t... configure: error: in
> `/home/cec/.SW/VIM/Merc/vim/src':
> configure: error: cannot compute sizeof (off_t)
>
> Here's some excerpts from config.log:  (a centos 5.2 system)
> ...
>
> configure:11492: checking size of off_t
> configure:11497: gcc -o conftest -g -O2   -Wl,-E
> -Wl,-rpath,/usr/lib64/perl5/5.8.8/x86_64-linux-thread-multi/CORE
> -L/usr/local/lib conftest.c -lm -lncurses -lelf -lnsl  -lselinux  -lacl
> -lattr -lgpm >&5
> In file included from /usr/include/inttypes.h:28,
>                 from conftest.c:164:
> /usr/include/stdint.h:52: error: duplicate 'unsigned'
> /usr/include/stdint.h:52: error: two or more data types in declaration
> specifiers
> configure:11497: $? = 1
> configure: program exited with status 1

config.log should also show what the program was that it tried to
compile.  Posting that would be useful.

--
James
GPG Key: 1024D/61326D40 2003-09-02 James Vega <[hidden email]>

--
You received this message from the "vim_dev" maillist.
Do not top-post! Type your reply below the text you are replying to.
For more information, visit http://www.vim.org/maillist.php
Reply | Threaded
Open this post in threaded view
|

Re: vim 7.3

James Vega-3
In reply to this post by Charles Campbell
On Tue, Jun 22, 2010 at 3:03 PM, Charles Campbell
<[hidden email]> wrote:
> Vim 7.3 configured&compiled ok before this pull, and FWIW vim 7.2 configures
> and builds ok (with patches 1-444).

Do you know what revision you were at before you pulled?
configure/configure.in haven't changed the off_t check in almost a
month.

--
James
GPG Key: 1024D/61326D40 2003-09-02 James Vega <[hidden email]>

--
You received this message from the "vim_dev" maillist.
Do not top-post! Type your reply below the text you are replying to.
For more information, visit http://www.vim.org/maillist.php
Reply | Threaded
Open this post in threaded view
|

Re: vim 7.3

Charles Campbell
James Vega wrote:

> On Tue, Jun 22, 2010 at 3:03 PM, Charles Campbell
> <[hidden email]>  wrote:
>    
>> Vim 7.3 configured&compiled ok before this pull, and FWIW vim 7.2 configures
>> and builds ok (with patches 1-444).
>>      
> Do you know what revision you were at before you pulled?
> configure/configure.in haven't changed the off_t check in almost a
> month.
>    

Here's the entry in the log that you requested in the preceding email:

configure:11497: gcc -o conftest -g -O2   -L/usr/local/lib conftest.c
-lm -lncurses -lelf -lnsl  -lselinux -lacl -lattr -lgpm >&5

I don't know; however,

vim3 --version yields

VIM - Vi IMproved 7.3 BETA (2010 May 15, compiled May 19 2010 15:09:27)

and hg-log.vim has

Wed May 19 12:31:27 EDT 2010
requesting all changes
adding changesets
adding manifests
adding file changes
added 2188 changesets with 17715 changes to 2453 files (+2 heads)
updating to branch default
2283 files updated, 0 files merged, 0 files removed, 0 files unresolved

so its probably been a month since I did the update.  I was out of the
country for awhile.

Regards,
Chip Campbell

--
You received this message from the "vim_dev" maillist.
Do not top-post! Type your reply below the text you are replying to.
For more information, visit http://www.vim.org/maillist.php
Reply | Threaded
Open this post in threaded view
|

Re: vim 7.3

Tony Mechelynck
In reply to this post by Charles Campbell
On 22/06/10 21:03, Charles Campbell wrote:

> Hello,
>
> I just attempted to update my copy of vim 7.2 with:
>
> (date && hg pull -u) 2>&1 |tee -a ../hg-vim.log
>
> I followed that up with a gmake distclean
> and ./configure --with-features=huge --enable-gui=gtk2
> --enable-perlinterp --enable-pythoninterp --enable-gnome-check
> --enable-cscope
>
> Configure failed with the following message:
>
> checking size of off_t... configure: error: in
> `/home/cec/.SW/VIM/Merc/vim/src':
> configure: error: cannot compute sizeof (off_t)
>
> Here's some excerpts from config.log: (a centos 5.2 system)
>
> hostname = xorn.gsfc.nasa.gov
> uname -m = x86_64
> uname -r = 2.6.18-194.3.1.el5
> uname -s = Linux
> uname -v = #1 SMP Thu May 13 13:08:30 EDT 2010
>
> /usr/bin/uname -p = unknown
> /bin/uname -X = unknown
>
> /bin/arch = x86_64
> /usr/bin/arch -k = unknown
> /usr/convex/getsysinfo = unknown
> /usr/bin/hostinfo = unknown
> /bin/machine = unknown
> /usr/bin/oslevel = unknown
> /bin/universe = unknown
>
> ...
>
> configure:11492: checking size of off_t
> configure:11497: gcc -o conftest -g -O2 -Wl,-E
> -Wl,-rpath,/usr/lib64/perl5/5.8.8/x86_64-linux-thread-multi/CORE
> -L/usr/local/lib conftest.c -lm -lncurses -lelf -lnsl -lselinux -lacl
> -lattr -lgpm >&5
> In file included from /usr/include/inttypes.h:28,
> from conftest.c:164:
> /usr/include/stdint.h:52: error: duplicate 'unsigned'
> /usr/include/stdint.h:52: error: two or more data types in declaration
> specifiers
> configure:11497: $? = 1
> configure: program exited with status 1
>
> Here's an excerpt from /usr/include/stdint.h, starting at line 51:
>
> #ifndef __uint32_t_defined
> typedef unsigned int uint32_t;
> # define __uint32_t_defined
> #endif
>
> Vim 7.3 configured&compiled ok before this pull, and FWIW vim 7.2
> configures and builds ok (with patches 1-444).
>
> Regards,
> Chip Campbell
>

In my experience it is dangerous to run configure manually, because
sometimes (when there have been changes in the configure scripts IIUC)
make will itself launch configure while "rebuilding the makefiles". In
some cases this will even result in configure being run twice one after
another in a single make run. If make runs configure and doesn't know
your configure arguments, you'll lose them, or part of them.

My gvim 7.3a (very similar to yours: Linux 32bit Huge +gui_gtk2
+gui_gnome +cscope -mzscheme +perl +python +ruby +tcl) compiles with no
problem from the latest sources as of Tue 22 Jun 06:28:58 2010 +0002
(and auto/config.h defines SIZEOF_OFF_T as 8). I just ran "make
reconfig" to make sure, after checking that auto/configure (which, in my
shadowdir, is a softlink to ../../configure) pointed to something no
older than configure.in etc. See at
http://users.skynet.be/antoine.mechelynck/vim/compunix.htm how I set my
configure arguments in environment variables passed to make, so that
whenever configure is run, even "by surprise" by make, it always gets
the settings I chose.

Currently I compile Vim in a "shadow directory" built by "make shadow"
under the src/ directory and mentioned in .hgignore. I have replaced the
Makefile copy there by a softlink, since my build methods require no
featureset-choice changes to the Makefile. The script to set the
environment variables is in that shadow directory, so after updating the
repository I run the following in another more or less permanent shell
two directory levels down:

        (date && source mycfg && make) 2>&1 |tee -a make.log
and if successful
        (date && source mycfg && make install) 2>&1 |tee -a make.log

This "make" program is GNU make, and /usr/bin/gmake is here a softlink
to it.)


Now, about where this off_t definition comes from. Running cscope over
the Vim source, I find the following:

  1  76  src/os_amiga.h           <<off_t>>
         typedef long off_t
  2  67  src/os_msdos.h           <<off_t>>
         typedef long off_t
  3  62  src_os_win16.h           <<off_t>>
         typedef long off_t
  4  76  /usr/include/sys/stat.h  <<off_t>>
         typedef __off_t off_t
  5  78  /usr/include/sys/stat.h  <<off_t>>
         typedef __off64_t off_t
  6  88  /usr/include/sys/types.h <<off_t>>
         typedef __off_t off_t
  7  90  /usr/include/sys/types.h <<off_t>>
         typedef __off64_t off_t
  8 244  /usr/include/unistd.h    <<off_t>>
         typedef __off_t off_t
  9 246  /usr/include/unistd.h    <<off_t>>
         typedef __off64_t off_t

and I see that (not counting the first three occurrences which are
obviously for other platforms), all three of these include files are
included by os_unix.h at lines 28, 29 and 56.


Best regards,
Tony.
--
Ingrate, n.:
        A man who bites the hand that feeds him, and then complains of
indigestion.

--
You received this message from the "vim_dev" maillist.
Do not top-post! Type your reply below the text you are replying to.
For more information, visit http://www.vim.org/maillist.php
Reply | Threaded
Open this post in threaded view
|

Re: vim 7.3

Tony Mechelynck
On 22/06/10 23:06, Tony Mechelynck wrote:

> On 22/06/10 21:03, Charles Campbell wrote:
>> Hello,
>>
>> I just attempted to update my copy of vim 7.2 with:
>>
>> (date && hg pull -u) 2>&1 |tee -a ../hg-vim.log
>>
>> I followed that up with a gmake distclean
>> and ./configure --with-features=huge --enable-gui=gtk2
>> --enable-perlinterp --enable-pythoninterp --enable-gnome-check
>> --enable-cscope
>>
>> Configure failed with the following message:
>>
>> checking size of off_t... configure: error: in
>> `/home/cec/.SW/VIM/Merc/vim/src':
>> configure: error: cannot compute sizeof (off_t)
>>
>> Here's some excerpts from config.log: (a centos 5.2 system)
>>
>> hostname = xorn.gsfc.nasa.gov
>> uname -m = x86_64
>> uname -r = 2.6.18-194.3.1.el5
>> uname -s = Linux
>> uname -v = #1 SMP Thu May 13 13:08:30 EDT 2010
>>
>> /usr/bin/uname -p = unknown
>> /bin/uname -X = unknown
>>
>> /bin/arch = x86_64
>> /usr/bin/arch -k = unknown
>> /usr/convex/getsysinfo = unknown
>> /usr/bin/hostinfo = unknown
>> /bin/machine = unknown
>> /usr/bin/oslevel = unknown
>> /bin/universe = unknown
>>
>> ...
>>
>> configure:11492: checking size of off_t
>> configure:11497: gcc -o conftest -g -O2 -Wl,-E
>> -Wl,-rpath,/usr/lib64/perl5/5.8.8/x86_64-linux-thread-multi/CORE
>> -L/usr/local/lib conftest.c -lm -lncurses -lelf -lnsl -lselinux -lacl
>> -lattr -lgpm >&5
>> In file included from /usr/include/inttypes.h:28,
>> from conftest.c:164:
>> /usr/include/stdint.h:52: error: duplicate 'unsigned'
>> /usr/include/stdint.h:52: error: two or more data types in declaration
>> specifiers
>> configure:11497: $? = 1
>> configure: program exited with status 1
>>
>> Here's an excerpt from /usr/include/stdint.h, starting at line 51:
>>
>> #ifndef __uint32_t_defined
>> typedef unsigned int uint32_t;
>> # define __uint32_t_defined
>> #endif
>>
>> Vim 7.3 configured&compiled ok before this pull, and FWIW vim 7.2
>> configures and builds ok (with patches 1-444).
>>
>> Regards,
>> Chip Campbell
>>
>
> In my experience it is dangerous to run configure manually, because
> sometimes (when there have been changes in the configure scripts IIUC)
> make will itself launch configure while "rebuilding the makefiles". In
> some cases this will even result in configure being run twice one after
> another in a single make run. If make runs configure and doesn't know
> your configure arguments, you'll lose them, or part of them.
>
> My gvim 7.3a (very similar to yours: Linux 32bit Huge +gui_gtk2
> +gui_gnome +cscope -mzscheme +perl +python +ruby +tcl) compiles with no
> problem from the latest sources as of Tue 22 Jun 06:28:58 2010 +0002
> (and auto/config.h defines SIZEOF_OFF_T as 8). I just ran "make
> reconfig" to make sure, after checking that auto/configure (which, in my
> shadowdir, is a softlink to ../../configure) pointed to something no
> older than configure.in etc. See at
> http://users.skynet.be/antoine.mechelynck/vim/compunix.htm how I set my
> configure arguments in environment variables passed to make, so that
> whenever configure is run, even "by surprise" by make, it always gets
> the settings I chose.
>
> Currently I compile Vim in a "shadow directory" built by "make shadow"
> under the src/ directory and mentioned in .hgignore. I have replaced the
> Makefile copy there by a softlink, since my build methods require no
> featureset-choice changes to the Makefile. The script to set the
> environment variables is in that shadow directory, so after updating the
> repository I run the following in another more or less permanent shell
> two directory levels down:
>
> (date && source mycfg && make) 2>&1 |tee -a make.log
> and if successful
> (date && source mycfg && make install) 2>&1 |tee -a make.log
>
> This "make" program is GNU make, and /usr/bin/gmake is here a softlink
> to it.)
>
>
> Now, about where this off_t definition comes from. Running cscope over
> the Vim source, I find the following:
>
> 1 76 src/os_amiga.h <<off_t>>
> typedef long off_t
> 2 67 src/os_msdos.h <<off_t>>
> typedef long off_t
> 3 62 src_os_win16.h <<off_t>>
> typedef long off_t
> 4 76 /usr/include/sys/stat.h <<off_t>>
> typedef __off_t off_t
> 5 78 /usr/include/sys/stat.h <<off_t>>
> typedef __off64_t off_t
> 6 88 /usr/include/sys/types.h <<off_t>>
> typedef __off_t off_t
> 7 90 /usr/include/sys/types.h <<off_t>>
> typedef __off64_t off_t
> 8 244 /usr/include/unistd.h <<off_t>>
> typedef __off_t off_t
> 9 246 /usr/include/unistd.h <<off_t>>
> typedef __off64_t off_t
>
> and I see that (not counting the first three occurrences which are
> obviously for other platforms), all three of these include files are
> included by os_unix.h at lines 28, 29 and 56.
>
>
> Best regards,
> Tony.

P.S. I have the same /usr/include/stdint.h as you do (at least, at lines
51-54).

Now let's try to find out all the places where that uint32_t symbol is
defined or used:

used 44 times in /usr/include/netinet/in.h
defined once in /usr/include/stdint.h
used once in /usr/include/bits/netdb.h
used 3 times in vim.h as follows:
vim.h|54| <<UINT32_TYPEDEF>> #define UINT32_TYPEDEF uint32_t
vim.h|58| <<defined>> #if defined(uint32_t)
vim.h|59| <<UINT32_TYPEDEF>> #define UINT32_TYPEDEF uint32_t
used twice in /usr/include/netdb.h

I think the relevant one is the one at vim.h:54


Best regards,
Tony.
--
May a Misguided Platypus lay its Eggs in your Jockey Shorts

--
You received this message from the "vim_dev" maillist.
Do not top-post! Type your reply below the text you are replying to.
For more information, visit http://www.vim.org/maillist.php
Reply | Threaded
Open this post in threaded view
|

Re: vim 7.3

James Vega-3
In reply to this post by Charles Campbell
On Tue, Jun 22, 2010 at 3:03 PM, Charles Campbell
<[hidden email]> wrote:

> Hello,
>
> I just attempted to update my copy of vim 7.2 with:
>
>      (date && hg pull -u) 2>&1 |tee -a ../hg-vim.log
>
> I followed that up with a   gmake distclean
> and                                     ./configure --with-features=huge
> --enable-gui=gtk2 --enable-perlinterp --enable-pythoninterp
> --enable-gnome-check --enable-cscope
>
> Configure failed with the following message:
>
> checking size of off_t... configure: error: in
> `/home/cec/.SW/VIM/Merc/vim/src':
> configure: error: cannot compute sizeof (off_t)
>
> Here's some excerpts from config.log:  (a centos 5.2 system)

Since CentOS 5.2 uses libraries a bit older than the systems I normally
test on, I setup a VM to see if it was something specific to the
library/toolchain versions in 5.2.  I installed CentOS, made a fresh
clone of the Vim repository, updated to the vim73 branch, and ran the
configure line you gave above.  Everything ran fine for me and I was
able to build Vim.

Given that, I'm not sure why you're seeing different results, unless the
different architecture is affecting things somehow (your 64-bit x86_64
vs. my 32-bit x86).

--
James
GPG Key: 1024D/61326D40 2003-09-02 James Vega <[hidden email]>

--
You received this message from the "vim_dev" maillist.
Do not top-post! Type your reply below the text you are replying to.
For more information, visit http://www.vim.org/maillist.php
Reply | Threaded
Open this post in threaded view
|

Re: vim 7.3

Charles Campbell
James Vega wrote:

> On Tue, Jun 22, 2010 at 3:03 PM, Charles Campbell
> <[hidden email]>  wrote:
>    
>> Hello,
>>
>> I just attempted to update my copy of vim 7.2 with:
>>
>>       (date&&  hg pull -u) 2>&1 |tee -a ../hg-vim.log
>>
>> I followed that up with a   gmake distclean
>> and                                     ./configure --with-features=huge
>> --enable-gui=gtk2 --enable-perlinterp --enable-pythoninterp
>> --enable-gnome-check --enable-cscope
>>
>> Configure failed with the following message:
>>
>> checking size of off_t... configure: error: in
>> `/home/cec/.SW/VIM/Merc/vim/src':
>> configure: error: cannot compute sizeof (off_t)
>>
>> Here's some excerpts from config.log:  (a centos 5.2 system)
>>      
> Since CentOS 5.2 uses libraries a bit older than the systems I normally
> test on, I setup a VM to see if it was something specific to the
> library/toolchain versions in 5.2.  I installed CentOS, made a fresh
> clone of the Vim repository, updated to the vim73 branch, and ran the
> configure line you gave above.  Everything ran fine for me and I was
> able to build Vim.
>
> Given that, I'm not sure why you're seeing different results, unless the
> different architecture is affecting things somehow (your 64-bit x86_64
> vs. my 32-bit x86).
>    

Thank you for going to all that trouble!  I started getting additional
messages (like "configure: error: `CC' was not set in the previous run")
and make distclean wouldn't complete.  So, I applied "/bin/rm -rf" to
the vim directory, and pulled a new copy.  This time it configured and
compiled without problems.  I suppose I should've tried that earlier.

Regards,
Chip Campbell

--
You received this message from the "vim_dev" maillist.
Do not top-post! Type your reply below the text you are replying to.
For more information, visit http://www.vim.org/maillist.php