file-3.36 [lfs-packages-cvs-20010926-0930.tar]

Erika Pacholleck pchllck at
Mon Oct 15 02:51:07 PDT 2001

[14.10.01 11:39 +0100] Chris Lingard <-- :
> You can always revert to an older version of file;  but they are all wrong.
> The reason is that it is looking in the wrong place for uint8 ...  What the 
> configure stuff does it to make up its own definitions;  this is why it 
> cannot detect them; but can calculate their size.
> Gerard's patch works fine.  It puts in the header file stdint.h so that the 
> correct definitions are found in the compile phase.

Hi Chris,
ok, I understand that finally configure results seems to be less
important. And I did a comparision compiling with the 1.4-p1 version
for 3.35 and 3.36 on my "now-lfs". 

From the Makefile line COMPILING= $(CC) $(DEFS)<-- =-DHAVE_CONFIG_H
I guess that config.h will be used during compile. So I had a look on
what happens on my "now-lfs": in config.h all uintxx_t's are #defined
and all sizes have their correct values defined. Make just runs fine.

Now the 1.5/3.36 combination configures the config.h exactely as the
old compare-config.h. But at the start of make all these aclocal,
automake, autoconf mess it all up: config.h has undefined 8,16,32
uint's and their values are 0. And exactely the line we added is causing
the warning. I don't think that adding the stdint.h is supposed to
create an empty declaration.

Ok, one could say, it is just a warning, forget it; file seems to work.
But I can't help thinking that there is something going wrong here.
For now I will have a bear on it, laugh about all those unnecessary
thoughts and leave it as it is. Cheers ;)
Erika Pacholleck <pchllck at> @@ -18,9 +18,12 @@ -p+pa -h+ho -c+ec
Unsubscribe: send email to listar at
and put 'unsubscribe blfs-support' in the subject header of the message

More information about the blfs-support mailing list