Hi Tony.
On Fri, Nov 27, 2009 at 03:50:30PM +0000, Tony Finch wrote:
> Fix handling of input files (e.g. with no newline at EOF) that could
> make unifdef get into an unexpected state and call abort().
>
> The new -B option compresses blank lines around a deleted section
> so that blank lines around "paragraphs" of code don't get doubled.
>
> The evaluator can now handle macros with arguments, and unbracketed
> arguments to the "defined" operator.
Can you confirm that this does not regress with respect to the changes
that Russell King implemented in the commit:
eedc9d83eaab2d35fb9dd1ec25b765dec964e26c
("kbuild: fix headers_exports with boolean expression")
I long time ago started a small project to do a dedicated
implementation of unifdef solely for use by the kernel.
Today we do some preprocessing in perl - and then
we pass the result to unifdef file by file.
A dedicated program could do this much faster -
and this would be a bigger incentive for the kernel supplied
tools to actually use the exported headers.
But I never got far with it and I think the code was
lost when I changed computer some time ago.
Sam
On Fri, 27 Nov 2009, Sam Ravnborg wrote:
>
> Can you confirm that this does not regress with respect to the changes
> that Russell King implemented in the commit:
> eedc9d83eaab2d35fb9dd1ec25b765dec964e26c
> ("kbuild: fix headers_exports with boolean expression")
Yes. I ran the following commands:
git checkout master
gmake distclean
gmake headers_install ARCH=i386 INSTALL_HDR_PATH=../include-master
git checkout fanf-unifdef
gmake distclean
gmake headers_install ARCH=i386 INSTALL_HDR_PATH=../include-fanf
diff -Nru -x ..install.cmd ../include-master ../include-fanf
and diff produced no output.
I didn't mention Russell's change in the commit message since its
functionality isn't affected, although the code is different - Ben
Hutchings contributed it to me in February/March 2008.
Tony.
--
f.anthony.n.finch <[email protected]> http://dotat.at/
GERMAN BIGHT HUMBER: SOUTHWEST 5 TO 7. MODERATE OR ROUGH. SQUALLY SHOWERS.
MODERATE OR GOOD.
On Fri, Nov 27, 2009 at 07:15:01PM +0000, Tony Finch wrote:
> On Fri, 27 Nov 2009, Sam Ravnborg wrote:
> >
> > Can you confirm that this does not regress with respect to the changes
> > that Russell King implemented in the commit:
> > eedc9d83eaab2d35fb9dd1ec25b765dec964e26c
> > ("kbuild: fix headers_exports with boolean expression")
>
> Yes. I ran the following commands:
>
> git checkout master
> gmake distclean
> gmake headers_install ARCH=i386 INSTALL_HDR_PATH=../include-master
> git checkout fanf-unifdef
> gmake distclean
> gmake headers_install ARCH=i386 INSTALL_HDR_PATH=../include-fanf
> diff -Nru -x ..install.cmd ../include-master ../include-fanf
>
> and diff produced no output.
Good.
Michael - you can add my:
Acked-by: Sam Ravnborg <[email protected]>
Sam
On 27.11.2009 19:17, Sam Ravnborg wrote:
> On Fri, Nov 27, 2009 at 03:50:30PM +0000, Tony Finch wrote:
>> Fix handling of input files (e.g. with no newline at EOF) that could
>> make unifdef get into an unexpected state and call abort().
>>
>> The new -B option compresses blank lines around a deleted section
>> so that blank lines around "paragraphs" of code don't get doubled.
>>
>> The evaluator can now handle macros with arguments, and unbracketed
>> arguments to the "defined" operator.
Hi Tony,
I don't have the patch and I can't find it in the archives either. Could
you please resend it?
Thanks!
Michal