[Buildroot] [PATCH] ncurses: fix wide-char enabled compilation for noMMU targets

[Buildroot] [PATCH] ncurses: fix wide-char enabled compilation for noMMU targets

[Waldemar Brodkorb]

For noMMU targets -D_XOPEN_SOURCE_EXTENDED must be explicitely
passed to the preprocessor to allow cchar_t usage.

Fixes:
http://autobuild.buildroot.net/results/5bb34ff490c70eea5e4fb497e5228ca1319fffdc/
http://autobuild.buildroot.net/results/8ba1410ed3ffb4954ccc4b7c3996d1839d677bef/
http://autobuild.buildroot.net/results/26ee52ad549b7ef75c9ce4b2eae94f9312cea775/

Signed-off-by: Waldemar Brodkorb <wbx@openadk.org>
---
 package/ncurses/Config.in  |    2 --
 package/ncurses/ncurses.mk |    4 ++++
 2 files changed, 4 insertions(+), 2 deletions(-)

diff --git a/package/ncurses/Config.in b/package/ncurses/Config.in
index 44713f9..8c8acde 100644
--- a/package/ncurses/Config.in
+++ b/package/ncurses/Config.in
@@ -13,8 +13,6 @@ if BR2_PACKAGE_NCURSES
 config BR2_PACKAGE_NCURSES_WCHAR
 	bool "enable wide char support"
 	depends on BR2_USE_WCHAR
-	# Build broken @ curses.priv.h with bad declarations
-	depends on !(BR2_bfin && BR2_BINFMT_FLAT)
 	help
 	  Enable wide char & UTF-8 support in ncurses libraries
 
diff --git a/package/ncurses/ncurses.mk b/package/ncurses/ncurses.mk
index bef57c5..412f548 100644
--- a/package/ncurses/ncurses.mk
+++ b/package/ncurses/ncurses.mk
@@ -57,8 +57,12 @@ NCURSES_LIBS-$(BR2_PACKAGE_NCURSES_TARGET_FORM) += form
 
 ifeq ($(BR2_PACKAGE_NCURSES_WCHAR),y)
 NCURSES_CONF_OPTS += --enable-widec
+# for noMMU we need to set it explicitely, fixes cchar_t problem
+NCURSES_CPPFLAGS += -D_XOPEN_SOURCE_EXTENDED
 NCURSES_LIB_SUFFIX = w
 
+NCURSES_CONF_ENV += CPPFLAGS="$(NCURSES_CPPFLAGS)"
+
 define NCURSES_LINK_LIBS_STATIC
 	for lib in $(NCURSES_LIBS-y:%=lib%); do \
 		ln -sf $${lib}$(NCURSES_LIB_SUFFIX).a \
-- 
1.7.10.4

[Buildroot] [PATCH] ncurses: fix wide-char enabled compilation for noMMU targets

[Khem Raj]

> On Jul 31, 2016, at 1:28 AM, Waldemar Brodkorb <wbx@openadk.org> wrote:
> 
> For noMMU targets -D_XOPEN_SOURCE_EXTENDED must be explicitely
> passed to the preprocessor to allow cchar_t usage.
> 
> Fixes:
> http://autobuild.buildroot.net/results/5bb34ff490c70eea5e4fb497e5228ca1319fffdc/
> http://autobuild.buildroot.net/results/8ba1410ed3ffb4954ccc4b7c3996d1839d677bef/
> http://autobuild.buildroot.net/results/26ee52ad549b7ef75c9ce4b2eae94f9312cea775/
> 
> Signed-off-by: Waldemar Brodkorb <wbx@openadk.org>
> ---
> package/ncurses/Config.in  |    2 --
> package/ncurses/ncurses.mk |    4 ++++
> 2 files changed, 4 insertions(+), 2 deletions(-)
> 
> diff --git a/package/ncurses/Config.in b/package/ncurses/Config.in
> index 44713f9..8c8acde 100644
> --- a/package/ncurses/Config.in
> +++ b/package/ncurses/Config.in
> @@ -13,8 +13,6 @@ if BR2_PACKAGE_NCURSES
> config BR2_PACKAGE_NCURSES_WCHAR
> 	bool "enable wide char support"
> 	depends on BR2_USE_WCHAR
> -	# Build broken @ curses.priv.h with bad declarations
> -	depends on !(BR2_bfin && BR2_BINFMT_FLAT)
> 	help
> 	  Enable wide char & UTF-8 support in ncurses libraries
> 
> diff --git a/package/ncurses/ncurses.mk b/package/ncurses/ncurses.mk
> index bef57c5..412f548 100644
> --- a/package/ncurses/ncurses.mk
> +++ b/package/ncurses/ncurses.mk
> @@ -57,8 +57,12 @@ NCURSES_LIBS-$(BR2_PACKAGE_NCURSES_TARGET_FORM) += form
> 
> ifeq ($(BR2_PACKAGE_NCURSES_WCHAR),y)
> NCURSES_CONF_OPTS += --enable-widec
> +# for noMMU we need to set it explicitely, fixes cchar_t problem
> +NCURSES_CPPFLAGS += -D_XOPEN_SOURCE_EXTENDED

May be you need to define NCURSES_WIDECHAR=1 instead

> NCURSES_LIB_SUFFIX = w
> 
> +NCURSES_CONF_ENV += CPPFLAGS="$(NCURSES_CPPFLAGS)"
> +
> define NCURSES_LINK_LIBS_STATIC
> 	for lib in $(NCURSES_LIBS-y:%=lib%); do \
> 		ln -sf $${lib}$(NCURSES_LIB_SUFFIX).a \
> --
> 1.7.10.4
> 
> _______________________________________________
> buildroot mailing list
> buildroot at busybox.net
> http://lists.busybox.net/mailman/listinfo/buildroot

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 204 bytes
Desc: Message signed with OpenPGP using GPGMail
URL: <http://lists.busybox.net/pipermail/buildroot/attachments/20160803/f5361583/attachment.asc>

[Buildroot] [PATCH] ncurses: fix wide-char enabled compilation for noMMU targets

[Thomas Petazzoni]

Hello,

On Wed, 3 Aug 2016 23:47:51 -0700, Khem Raj wrote:

> > ifeq ($(BR2_PACKAGE_NCURSES_WCHAR),y)
> > NCURSES_CONF_OPTS += --enable-widec
> > +# for noMMU we need to set it explicitely, fixes cchar_t problem
> > +NCURSES_CPPFLAGS += -D_XOPEN_SOURCE_EXTENDED  
> 
> May be you need to define NCURSES_WIDECHAR=1 instead

Cannot work (at least without patching) : it gets undef'ed by ncurses:

#undef NCURSES_WIDECHAR
#if defined(_XOPEN_SOURCE_EXTENDED) || defined(_XPG5)
#define NCURSES_WIDECHAR
#endif

But this test from ncurses is weird. My understanding of the feature
macros like _XOPEN_SOURCE_EXTENDED is that they should be *defined* by
the program/library that wants to use the specific features hidden
behind this macro. This is what
https://www.gnu.org/software/libc/manual/html_node/Feature-Test-Macros.html
explains.

So why is ncurses *testing* this?

Thomas
-- 
Thomas Petazzoni, CTO, Free Electrons
Embedded Linux and Kernel engineering
http://free-electrons.com

[Buildroot] [PATCH] ncurses: fix wide-char enabled compilation for noMMU targets

[Yann E. MORIN]

Waldemar, All,

On 2016-07-31 10:28 +0200, Waldemar Brodkorb spake thusly:
> For noMMU targets -D_XOPEN_SOURCE_EXTENDED must be explicitely
> passed to the preprocessor to allow cchar_t usage.
> 
> Fixes:
> http://autobuild.buildroot.net/results/5bb34ff490c70eea5e4fb497e5228ca1319fffdc/
> http://autobuild.buildroot.net/results/8ba1410ed3ffb4954ccc4b7c3996d1839d677bef/
> http://autobuild.buildroot.net/results/26ee52ad549b7ef75c9ce4b2eae94f9312cea775/
> 
> Signed-off-by: Waldemar Brodkorb <wbx@openadk.org>
> ---
>  package/ncurses/Config.in  |    2 --
>  package/ncurses/ncurses.mk |    4 ++++
>  2 files changed, 4 insertions(+), 2 deletions(-)
> 
> diff --git a/package/ncurses/Config.in b/package/ncurses/Config.in
> index 44713f9..8c8acde 100644
> --- a/package/ncurses/Config.in
> +++ b/package/ncurses/Config.in
> @@ -13,8 +13,6 @@ if BR2_PACKAGE_NCURSES
>  config BR2_PACKAGE_NCURSES_WCHAR
>  	bool "enable wide char support"
>  	depends on BR2_USE_WCHAR
> -	# Build broken @ curses.priv.h with bad declarations
> -	depends on !(BR2_bfin && BR2_BINFMT_FLAT)
>  	help
>  	  Enable wide char & UTF-8 support in ncurses libraries
>  
> diff --git a/package/ncurses/ncurses.mk b/package/ncurses/ncurses.mk
> index bef57c5..412f548 100644
> --- a/package/ncurses/ncurses.mk
> +++ b/package/ncurses/ncurses.mk
> @@ -57,8 +57,12 @@ NCURSES_LIBS-$(BR2_PACKAGE_NCURSES_TARGET_FORM) += form
>  
>  ifeq ($(BR2_PACKAGE_NCURSES_WCHAR),y)
>  NCURSES_CONF_OPTS += --enable-widec
> +# for noMMU we need to set it explicitely, fixes cchar_t problem
> +NCURSES_CPPFLAGS += -D_XOPEN_SOURCE_EXTENDED

As noticed by Thomas, I don't think this is the correct solution.

If you look at configure.in, you'll see they have a test to decide
whether _XOPEN_SOURCE_EXTENDED is needed or not. For targets with an
MMU, this test is checked, while it is not for a noMMU target. (Yes, for
an MMU platform, the test concludes that defining it is not needed.)

I think the correct solution would be to see why this test does not pass
for noMMU.

Adn the reason it does not is that it decides to add -D_XOPEN_SOURCE=500
to CPPFLAGS when calling CF_XOPEN_SOURCE(500) (line 697 in configure.in)

The function checks for ${host_os} for various well-known values. One
interesting well-known value is (sxerpt from aclocal.m4):

    6504 case $host_os in #(vi
    [...]
    6534 linux*|gnu*|mint*|k*bsd*-gnu) #(vi
    6535     CF_GNU_SOURCE
    6536     ;;
    [...]
    6561 *)
    6562     AC_CACHE_CHECK(if we should define _XOPEN_SOURCE,cf_cv_xopen_source,[
    [...]
    6587 esac

See the value on line 6534? It is only about 'linux*'. But then when you
look at how we call ./configure:

    ./configure \
        --target=m68k-buildroot-uclinux-uclibc \
        --host=m68k-buildroot-uclinux-uclibc \
        [...]

So, host_os=uclinux-uclibc and then it does not match 'linux*' and then
CF_XOPEN_SOURCE falls back to the generic case, which checks wrong for
uclinux.

So, the coirrect solution would seem to be to fix CF_XOPEN_SOURCE to
also include 'uclinux*' in the case, above.

And indeed, doing so does fix the build. ;-)

I'll send a patch shortly (after I clean up all my dirty traces in
configure...). In the meantime, I've marked your patch as rejected in
patchwork.

Thanks for triggering this investigation on my side! ;-)

Regards,
Yann E. MORIN.

>  NCURSES_LIB_SUFFIX = w
>  
> +NCURSES_CONF_ENV += CPPFLAGS="$(NCURSES_CPPFLAGS)"
> +
>  define NCURSES_LINK_LIBS_STATIC
>  	for lib in $(NCURSES_LIBS-y:%=lib%); do \
>  		ln -sf $${lib}$(NCURSES_LIB_SUFFIX).a \
> -- 
> 1.7.10.4
> 
> _______________________________________________
> buildroot mailing list
> buildroot at busybox.net
> http://lists.busybox.net/mailman/listinfo/buildroot

-- 
.-----------------.--------------------.------------------.--------------------.
|  Yann E. MORIN  | Real-Time Embedded | /"\ ASCII RIBBON | Erics' conspiracy: |
| +33 662 376 056 | Software  Designer | \ / CAMPAIGN     |  ___               |
| +33 223 225 172 `------------.-------:  X  AGAINST      |  \e/  There is no  |
| http://ymorin.is-a-geek.org/ | _/*\_ | / \ HTML MAIL    |   v   conspiracy.  |
'------------------------------^-------^------------------^--------------------'