| Message ID | 20150224235024.844FBE0447@smtp.hushmail.com |
|---|---|
| State | Rejected |
| Headers | show |
Justin, All, On 2015-02-25 01:50 +0200, justin.swartz@nym.hush.com spake thusly: > Fixes: http://article.gmane.org/gmane.comp.lib.uclibc.buildroot/65165 > > Credit to Yann E. MORIN for discovering and describing a fix for the > HAVE_DOCBOOK bug in "configure.ac", as well as the > "--disable-gtk-doc-pdf" and "--disable-gtk-doc-html" additions to > LIBGTK2_CONF_OPTS. > > Signed-off-by: Justin Swartz <justin.swartz@nym.hush.com> Tested-by: "Yann E. MORIN" <yann.morin.1998@free.fr> Acked-by: "Yann E. MORIN" <yann.morin.1998@free.fr> > --- > Changes v1 -> v2: > - Commit log added to the head of "0002-configure-have_docbook.patch" > - Moved "--disable-gtk-doc-pdf" and "--disable-gtk-doc-html" to the > end of LIBGTK2_CONF_OPTS > --- > package/libgtk2/0002-configure-have_docbook.patch | 15 +++++++++++++++ > package/libgtk2/libgtk2.mk | 9 +++++++-- > 2 files changed, 22 insertions(+), 2 deletions(-) > create mode 100644 package/libgtk2/0002-configure-have_docbook.patch > > diff --git a/package/libgtk2/0002-configure-have_docbook.patch b/package/libgtk2/0002-configure-have_docbook.patch > new file mode 100644 > index 0000000..aece8f1 > --- /dev/null > +++ b/package/libgtk2/0002-configure-have_docbook.patch > @@ -0,0 +1,15 @@ > +Improvement of the HAVE_DOCBOOK test criteria. > + > +Signed-off-by: Justin Swartz <justin.swartz@nym.hush.com> I guess I can add my own SoB, as I was involved with that patch. ;-) Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr> Regards, Yann E. MORIN. > + > +--- libgtk2-2.24.18/configure.ac.orig 2015-02-24 00:28:18.128745222 +0200 > ++++ libgtk2-2.24.18/configure.ac 2015-02-24 00:29:10.739087501 +0200 > +@@ -1612,7 +1612,7 @@ > + GTK_DOC_CHECK([1.11]) > + > + AC_CHECK_PROG(DB2HTML, db2html, true, false) > +-AM_CONDITIONAL(HAVE_DOCBOOK, $DB2HTML) > ++AM_CONDITIONAL([HAVE_DOCBOOK], [test ! "$DB2HTML" = "no"]) > + > + AC_ARG_ENABLE(man, > + [AC_HELP_STRING([--enable-man], > diff --git a/package/libgtk2/libgtk2.mk b/package/libgtk2/libgtk2.mk > index dfc1f5b..d468117 100644 > --- a/package/libgtk2/libgtk2.mk > +++ b/package/libgtk2/libgtk2.mk > @@ -69,9 +69,14 @@ LIBGTK2_CONF_ENV = \ > ac_cv_path_GTK_UPDATE_ICON_CACHE=$(HOST_DIR)/usr/bin/gtk-update-icon-cache \ > ac_cv_path_GDK_PIXBUF_CSOURCE=$(HOST_DIR)/usr/bin/gdk-pixbuf-csource \ > ac_cv_prog_F77=no \ > - ac_cv_path_CUPS_CONFIG=no > + ac_cv_path_CUPS_CONFIG=no \ > + ac_cv_prog_DB2HTML=no > > -LIBGTK2_CONF_OPTS = --disable-glibtest --enable-explicit-deps=no > +LIBGTK2_CONF_OPTS = \ > + --disable-glibtest \ > + --enable-explicit-deps=no \ > + --disable-gtk-doc-pdf \ > + --disable-gtk-doc-html > > LIBGTK2_DEPENDENCIES = host-pkgconf host-libgtk2 libglib2 cairo pango atk gdk-pixbuf > > -- > 1.7.12.1 >
>>>>> "justin" == justin swartz <justin.swartz@nym.hush.com> writes: > Fixes: http://article.gmane.org/gmane.comp.lib.uclibc.buildroot/65165 > Credit to Yann E. MORIN for discovering and describing a fix for the > HAVE_DOCBOOK bug in "configure.ac", as well as the > "--disable-gtk-doc-pdf" and "--disable-gtk-doc-html" additions to > LIBGTK2_CONF_OPTS. > Signed-off-by: Justin Swartz <justin.swartz@nym.hush.com> > --- > Changes v1 -> v2: > - Commit log added to the head of "0002-configure-have_docbook.patch" > - Moved "--disable-gtk-doc-pdf" and "--disable-gtk-doc-html" to the > end of LIBGTK2_CONF_OPTS > --- > package/libgtk2/0002-configure-have_docbook.patch | 15 +++++++++++++++ > package/libgtk2/libgtk2.mk | 9 +++++++-- > 2 files changed, 22 insertions(+), 2 deletions(-) > create mode 100644 package/libgtk2/0002-configure-have_docbook.patch > diff --git a/package/libgtk2/0002-configure-have_docbook.patch b/package/libgtk2/0002-configure-have_docbook.patch > new file mode 100644 > index 0000000..aece8f1 > --- /dev/null > +++ b/package/libgtk2/0002-configure-have_docbook.patch > @@ -0,0 +1,15 @@ > +Improvement of the HAVE_DOCBOOK test criteria. > + > +Signed-off-by: Justin Swartz <justin.swartz@nym.hush.com> > + > +--- libgtk2-2.24.18/configure.ac.orig 2015-02-24 00:28:18.128745222 +0200 > ++++ libgtk2-2.24.18/configure.ac 2015-02-24 00:29:10.739087501 +0200 > +@@ -1612,7 +1612,7 @@ > + GTK_DOC_CHECK([1.11]) > + > + AC_CHECK_PROG(DB2HTML, db2html, true, false) > +-AM_CONDITIONAL(HAVE_DOCBOOK, $DB2HTML) > ++AM_CONDITIONAL([HAVE_DOCBOOK], [test ! "$DB2HTML" = "no"]) Why not just pass DB2HTML=false in LIBGTK2_CONF_ENV? > + > + AC_ARG_ENABLE(man, > + [AC_HELP_STRING([--enable-man], > diff --git a/package/libgtk2/libgtk2.mk b/package/libgtk2/libgtk2.mk > index dfc1f5b..d468117 100644 > --- a/package/libgtk2/libgtk2.mk > +++ b/package/libgtk2/libgtk2.mk > @@ -69,9 +69,14 @@ LIBGTK2_CONF_ENV = \ > ac_cv_path_GTK_UPDATE_ICON_CACHE=$(HOST_DIR)/usr/bin/gtk-update-icon-cache \ > ac_cv_path_GDK_PIXBUF_CSOURCE=$(HOST_DIR)/usr/bin/gdk-pixbuf-csource \ > ac_cv_prog_F77=no \ > - ac_cv_path_CUPS_CONFIG=no > + ac_cv_path_CUPS_CONFIG=no \ > + ac_cv_prog_DB2HTML=no > -LIBGTK2_CONF_OPTS = --disable-glibtest --enable-explicit-deps=no > +LIBGTK2_CONF_OPTS = \ > + --disable-glibtest \ > + --enable-explicit-deps=no \ > + --disable-gtk-doc-pdf \ > + --disable-gtk-doc-html What does that buy us? gtk-doc is disabled unless --enable-gtk-doc is passed (which we don't), so these shouldn't change anything.
This should have been addressed to "peter@korsgaard.com", but my provider claims that the address is invalid. On 02/26/2015 at 1:17 AM, "Peter Korsgaard" <peter@korsgaard.com> wrote: >Why not just pass DB2HTML=false in LIBGTK2_CONF_ENV? > DB2HTML=false works perfectly, thanks for pointing that out! > > +LIBGTK2_CONF_OPTS = \ > > + --disable-glibtest \ > > + --enable-explicit-deps=no \ > > + --disable-gtk-doc-pdf \ > > + --disable-gtk-doc-html > >What does that buy us? gtk-doc is disabled unless --enable-gtk-doc >is passed (which we don't), so these shouldn't change anything. > I hadn't noticed that --disable-gtk-doc was implicitly passed in *_CONF_OPTS. Unfortunately there is still documention installed in to /usr/share/gtk-doc/html/, but it's eventually removed by target-finalize anyway. Regards Justin
Peter, All, On 2015-02-26 00:17 +0100, Peter Korsgaard spake thusly: > >>>>> "justin" == justin swartz <justin.swartz@nym.hush.com> writes: > > > Fixes: http://article.gmane.org/gmane.comp.lib.uclibc.buildroot/65165 > > Credit to Yann E. MORIN for discovering and describing a fix for the > > HAVE_DOCBOOK bug in "configure.ac", as well as the > > "--disable-gtk-doc-pdf" and "--disable-gtk-doc-html" additions to > > LIBGTK2_CONF_OPTS. > > > Signed-off-by: Justin Swartz <justin.swartz@nym.hush.com> > > > --- > > Changes v1 -> v2: > > - Commit log added to the head of "0002-configure-have_docbook.patch" > > - Moved "--disable-gtk-doc-pdf" and "--disable-gtk-doc-html" to the > > end of LIBGTK2_CONF_OPTS > > --- > > package/libgtk2/0002-configure-have_docbook.patch | 15 +++++++++++++++ > > package/libgtk2/libgtk2.mk | 9 +++++++-- > > 2 files changed, 22 insertions(+), 2 deletions(-) > > create mode 100644 package/libgtk2/0002-configure-have_docbook.patch > > > diff --git a/package/libgtk2/0002-configure-have_docbook.patch b/package/libgtk2/0002-configure-have_docbook.patch > > new file mode 100644 > > index 0000000..aece8f1 > > --- /dev/null > > +++ b/package/libgtk2/0002-configure-have_docbook.patch > > @@ -0,0 +1,15 @@ > > +Improvement of the HAVE_DOCBOOK test criteria. > > + > > +Signed-off-by: Justin Swartz <justin.swartz@nym.hush.com> > > + > > +--- libgtk2-2.24.18/configure.ac.orig 2015-02-24 00:28:18.128745222 +0200 > > ++++ libgtk2-2.24.18/configure.ac 2015-02-24 00:29:10.739087501 +0200 > > +@@ -1612,7 +1612,7 @@ > > + GTK_DOC_CHECK([1.11]) > > + > > + AC_CHECK_PROG(DB2HTML, db2html, true, false) > > +-AM_CONDITIONAL(HAVE_DOCBOOK, $DB2HTML) > > ++AM_CONDITIONAL([HAVE_DOCBOOK], [test ! "$DB2HTML" = "no"]) > > Why not just pass DB2HTML=false in LIBGTK2_CONF_ENV? Weird, I was under the impression I did try that and that it did not work... But Justin is reporting that it does work, so my memory are obviously flawed... > > + AC_ARG_ENABLE(man, > > + [AC_HELP_STRING([--enable-man], > > diff --git a/package/libgtk2/libgtk2.mk b/package/libgtk2/libgtk2.mk > > index dfc1f5b..d468117 100644 > > --- a/package/libgtk2/libgtk2.mk > > +++ b/package/libgtk2/libgtk2.mk > > @@ -69,9 +69,14 @@ LIBGTK2_CONF_ENV = \ > > ac_cv_path_GTK_UPDATE_ICON_CACHE=$(HOST_DIR)/usr/bin/gtk-update-icon-cache \ > > ac_cv_path_GDK_PIXBUF_CSOURCE=$(HOST_DIR)/usr/bin/gdk-pixbuf-csource \ > > ac_cv_prog_F77=no \ > > - ac_cv_path_CUPS_CONFIG=no > > + ac_cv_path_CUPS_CONFIG=no \ > > + ac_cv_prog_DB2HTML=no > > > -LIBGTK2_CONF_OPTS = --disable-glibtest --enable-explicit-deps=no > > +LIBGTK2_CONF_OPTS = \ > > + --disable-glibtest \ > > + --enable-explicit-deps=no \ > > + --disable-gtk-doc-pdf \ > > + --disable-gtk-doc-html > > What does that buy us? gtk-doc is disabled unless --enable-gtk-doc is > passed (which we don't), so these shouldn't change anything. Fact is, documentation is still generated even with --disable-gtk-doc. I'll do another test-build to double-check all that mess... Regards, Yann E. MORIN.
>>>>> "justin" == justin swartz <justin.swartz@nym.hush.com> writes: Hi, > This should have been addressed to "peter@korsgaard.com", but my > provider claims that the address is invalid. Really? What error message do you get exactly (perhaps privately to jacmet@sunsite.dk). > On 02/26/2015 at 1:17 AM, "Peter Korsgaard" <peter@korsgaard.com> wrote: >> Why not just pass DB2HTML=false in LIBGTK2_CONF_ENV? >> > DB2HTML=false works perfectly, thanks for pointing that out! Great. I've committed a fix to pass it. >> > +LIBGTK2_CONF_OPTS = \ >> > + --disable-glibtest \ >> > + --enable-explicit-deps=no \ >> > + --disable-gtk-doc-pdf \ >> > + --disable-gtk-doc-html >> >> What does that buy us? gtk-doc is disabled unless --enable-gtk-doc >> is passed (which we don't), so these shouldn't change anything. >> > I hadn't noticed that --disable-gtk-doc was implicitly passed in *_CONF_OPTS. That and the fact that it defaults to disabled: ./configure --help|grep gtk-doc --enable-gtk-doc use gtk-doc to build documentation [[default=no]] --enable-gtk-doc-html build documentation in html format [[default=yes]] --enable-gtk-doc-pdf build documentation in pdf format [[default=no]] > Unfortunately there is still documention installed in to /usr/share/gtk-doc/html/, > but it's eventually removed by target-finalize anyway. Yes, but the patch you sent didn't change anything about that, right?
Peter, All, On 2015-02-26 22:16 +0100, Peter Korsgaard spake thusly: > >>>>> "justin" == justin swartz <justin.swartz@nym.hush.com> writes: > > On 02/26/2015 at 1:17 AM, "Peter Korsgaard" <peter@korsgaard.com> wrote: > >> Why not just pass DB2HTML=false in LIBGTK2_CONF_ENV? > >> > > DB2HTML=false works perfectly, thanks for pointing that out! > > Great. I've committed a fix to pass it. Hmm... Didn't you see his v2? Regards, Yann E. MORIN.
>>>>> "Yann" == Yann E MORIN <yann.morin.1998@free.fr> writes: Hi, >> > + AC_CHECK_PROG(DB2HTML, db2html, true, false) >> > +-AM_CONDITIONAL(HAVE_DOCBOOK, $DB2HTML) >> > ++AM_CONDITIONAL([HAVE_DOCBOOK], [test ! "$DB2HTML" = "no"]) >> >> Why not just pass DB2HTML=false in LIBGTK2_CONF_ENV? > Weird, I was under the impression I did try that and that it did not > work... But Justin is reporting that it does work, so my memory are > obviously flawed... Yes, as far as I can see AC_CHECK_PROG directly uses the shell variable if defined: https://www.gnu.org/software/autoconf/manual/autoconf-2.68/html_node/Generic-Programs.html $DB2HTML directly expands to false, so AM_CONDITIONAL isn't true: https://www.gnu.org/software/automake/manual/html_node/Usage-of-Conditionals.html >> What does that buy us? gtk-doc is disabled unless --enable-gtk-doc >> is passed (which we don't), so these shouldn't change anything. > Fact is, documentation is still generated even with > --disable-gtk-doc. Really? That's not what I saw during my tests.
>>>>> "Yann" == Yann E MORIN <yann.morin.1998@free.fr> writes: > Peter, All, > On 2015-02-26 22:16 +0100, Peter Korsgaard spake thusly: >> >>>>> "justin" == justin swartz <justin.swartz@nym.hush.com> writes: >> > On 02/26/2015 at 1:17 AM, "Peter Korsgaard" <peter@korsgaard.com> wrote: >> >> Why not just pass DB2HTML=false in LIBGTK2_CONF_ENV? >> >> >> > DB2HTML=false works perfectly, thanks for pointing that out! >> >> Great. I've committed a fix to pass it. > Hmm... Didn't you see his v2? V3 you mean? Sorry, I didn't (slow mail).
diff --git a/package/libgtk2/0002-configure-have_docbook.patch b/package/libgtk2/0002-configure-have_docbook.patch new file mode 100644 index 0000000..aece8f1 --- /dev/null +++ b/package/libgtk2/0002-configure-have_docbook.patch @@ -0,0 +1,15 @@ +Improvement of the HAVE_DOCBOOK test criteria. + +Signed-off-by: Justin Swartz <justin.swartz@nym.hush.com> + +--- libgtk2-2.24.18/configure.ac.orig 2015-02-24 00:28:18.128745222 +0200 ++++ libgtk2-2.24.18/configure.ac 2015-02-24 00:29:10.739087501 +0200 +@@ -1612,7 +1612,7 @@ + GTK_DOC_CHECK([1.11]) + + AC_CHECK_PROG(DB2HTML, db2html, true, false) +-AM_CONDITIONAL(HAVE_DOCBOOK, $DB2HTML) ++AM_CONDITIONAL([HAVE_DOCBOOK], [test ! "$DB2HTML" = "no"]) + + AC_ARG_ENABLE(man, + [AC_HELP_STRING([--enable-man], diff --git a/package/libgtk2/libgtk2.mk b/package/libgtk2/libgtk2.mk index dfc1f5b..d468117 100644 --- a/package/libgtk2/libgtk2.mk +++ b/package/libgtk2/libgtk2.mk @@ -69,9 +69,14 @@ LIBGTK2_CONF_ENV = \ ac_cv_path_GTK_UPDATE_ICON_CACHE=$(HOST_DIR)/usr/bin/gtk-update-icon-cache \ ac_cv_path_GDK_PIXBUF_CSOURCE=$(HOST_DIR)/usr/bin/gdk-pixbuf-csource \ ac_cv_prog_F77=no \ - ac_cv_path_CUPS_CONFIG=no + ac_cv_path_CUPS_CONFIG=no \ + ac_cv_prog_DB2HTML=no -LIBGTK2_CONF_OPTS = --disable-glibtest --enable-explicit-deps=no +LIBGTK2_CONF_OPTS = \ + --disable-glibtest \ + --enable-explicit-deps=no \ + --disable-gtk-doc-pdf \ + --disable-gtk-doc-html LIBGTK2_DEPENDENCIES = host-pkgconf host-libgtk2 libglib2 cairo pango atk gdk-pixbuf
Fixes: http://article.gmane.org/gmane.comp.lib.uclibc.buildroot/65165 Credit to Yann E. MORIN for discovering and describing a fix for the HAVE_DOCBOOK bug in "configure.ac", as well as the "--disable-gtk-doc-pdf" and "--disable-gtk-doc-html" additions to LIBGTK2_CONF_OPTS. Signed-off-by: Justin Swartz <justin.swartz@nym.hush.com> --- Changes v1 -> v2: - Commit log added to the head of "0002-configure-have_docbook.patch" - Moved "--disable-gtk-doc-pdf" and "--disable-gtk-doc-html" to the end of LIBGTK2_CONF_OPTS --- package/libgtk2/0002-configure-have_docbook.patch | 15 +++++++++++++++ package/libgtk2/libgtk2.mk | 9 +++++++-- 2 files changed, 22 insertions(+), 2 deletions(-) create mode 100644 package/libgtk2/0002-configure-have_docbook.patch