[meta-java][PATCH 1/2] openjdk-8: fix MAKE detection patch

[meta-java][PATCH 1/2] openjdk-8: fix MAKE detection patch

[André Draszik]

From: André Draszik <andre.draszik@jci.com>

The patch had a few typos, leading to errors during ./configure
   ../jdk8u-4be07cb28b21/common/autoconf/configure: line 8408: test: too many arguments

Change-Id: I867eba7aae3390aa869e69c86f29e77b505043e7
---
 recipes-core/openjdk/patches-openjdk-8/dont-expect-fqpn-for-make.patch | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/recipes-core/openjdk/patches-openjdk-8/dont-expect-fqpn-for-make.patch b/recipes-core/openjdk/patches-openjdk-8/dont-expect-fqpn-for-make.patch
index 6454237..5192d1a 100644
--- a/recipes-core/openjdk/patches-openjdk-8/dont-expect-fqpn-for-make.patch
+++ b/recipes-core/openjdk/patches-openjdk-8/dont-expect-fqpn-for-make.patch
@@ -6,7 +6,7 @@
      # User has supplied a make, test it.
 -    if test ! -f "$MAKE"; then
 -      AC_MSG_ERROR([The specified make (by MAKE=$MAKE) is not found.])
-+    if test -a `dirname "$MAKE"` = "." -a ! -f "$MAKE"; then
++    if test `dirname "$MAKE"` = "." && ! test -f "$MAKE"; then
 +      AC_PATH_PROGS(CHECK_MAKE, $MAKE)
 +    else
 +      CHECK_MAKE="${MAKE}"
-- 
2.16.2

[meta-java][PATCH 2/2] openjdk-8: fix build with --as-needed host toolchains (Ubuntu 16.04)

[André Draszik]

From: André Draszik <andre.draszik@jci.com>

As per the commit message - build on hosts with --as-needed
toolchains (Ubuntu 16.04) using system provided zlib fails:

If the (host) toolchain has been configured to
unconditionally add --as-needed to the linker command line
then linking can fail when using system libraries.

The reason is that the order of command line arguments
becomes important with --as-needed and the JDK build system
places needed system libraries at the beginning of the
command line where it would normally place the object files
from its own bundled compiled version.

Having those system libraries early in the command line is
not useful, as they are discarded by the linker at that
point in time as it hasn't seen any reference to the
symbols provided yet.

As it seems a generic pattern in the makefiles here, just
place the $EXPECTED_OBJS early in the command line, before
any additional libraries, so as to fix this once and for
all.

Change-Id: I568067011bb8899700c63aa3ee9ca952045cce05
Signed-off-by: André Draszik <andre.draszik@jci.com>
---
 recipes-core/openjdk/openjdk-8-release-16xbyy.inc  |  1 +
 ...fix-build-on-as-needed-toolchains-generic.patch | 91 ++++++++++++++++++++++
 2 files changed, 92 insertions(+)
 create mode 100644 recipes-core/openjdk/patches-openjdk-8/0010-build-fix-build-on-as-needed-toolchains-generic.patch

diff --git a/recipes-core/openjdk/openjdk-8-release-16xbyy.inc b/recipes-core/openjdk/openjdk-8-release-16xbyy.inc
index ab72830..bd4a349 100644
--- a/recipes-core/openjdk/openjdk-8-release-16xbyy.inc
+++ b/recipes-core/openjdk/openjdk-8-release-16xbyy.inc
@@ -15,6 +15,7 @@ PATCHES_URI = "\
     file://0007-jdk-use-correct-include-for-poll.patch \
     file://0008-jdk-use-correct-include-for-signal.patch \
     file://0009-jdk-disable-backtrace-musl-build-fix.patch \
+    file://0010-build-fix-build-on-as-needed-toolchains-generic.patch \
 "
 # some patches extracted from http://cr.openjdk.java.net/~rkennke/shark-build-hotspot/webrev.01/hotspot.patch
 # reported via http://mail.openjdk.java.net/pipermail/build-dev/2015-January/013972.html
diff --git a/recipes-core/openjdk/patches-openjdk-8/0010-build-fix-build-on-as-needed-toolchains-generic.patch b/recipes-core/openjdk/patches-openjdk-8/0010-build-fix-build-on-as-needed-toolchains-generic.patch
new file mode 100644
index 0000000..2d02b7c
--- /dev/null
+++ b/recipes-core/openjdk/patches-openjdk-8/0010-build-fix-build-on-as-needed-toolchains-generic.patch
@@ -0,0 +1,91 @@
+From 84bcdb9cdab0e0be9cdfededfb518d3cea9009e3 Mon Sep 17 00:00:00 2001
+From: =?UTF-8?q?Andr=C3=A9=20Draszik?= <andre.draszik@jci.com>
+Date: Mon, 12 Mar 2018 15:40:58 +0000
+Subject: [PATCH] build: fix build on --as-needed toolchains (generic)
+MIME-Version: 1.0
+Content-Type: text/plain; charset=UTF-8
+Content-Transfer-Encoding: 8bit
+
+If the (host) toolchain has been configured to
+unconditionally add --as-needed to the linker command line
+then linking fails when using the system zlib:
+  + ...gcc -lz -L<sysroot>/usr/lib -L<sysroot>/lib \
+               -Wl,-rpath-link,<sysroot>/usr/lib -Wl,-rpath-link,<sysroot>/lib \
+               -Wl,-rpath,<sysroot>/usr/lib -Wl,-rpath,<sysroot>/lib \
+           -Wl,-O1 -Xlinker --hash-style=both -Xlinker -z -Xlinker defs -Xlinker -O1 \
+           -Xlinker --allow-shlib-undefined -Xlinker -soname=libunpack.so \
+           -Xlinker -z -Xlinker origin -Xlinker -rpath -Xlinker '$ORIGIN' \
+           -lc \
+           -Xlinker -version-script=<src>/jdk/make/mapfiles/libunpack/mapfile-vers-unpack200 \
+           -o $build/jdk/objs/unpackexe/unpack200 \
+           $build/jdk/objs/unpackexe/bands.o $build/jdk/objs/unpackexe/bytes.o \
+           $build/jdk/objs/unpackexe/coding.o $build/jdk/objs/unpackexe/main.o \
+           $build/jdk/objs/unpackexe/unpack.o $build/jdk/objs/unpackexe/utils.o \
+           $build/jdk/objs/unpackexe/zip.o -lstdc++
+  $build/jdk/objs/unpackexe/zip.o: In function `jar::deflate_bytes(bytes&, bytes&)':
+  $src/jdk/src/share/native/com/sun/java/util/jar/pack/zip.cpp:464: undefined reference to `deflateInit2_'
+  $src/jdk/src/share/native/com/sun/java/util/jar/pack/zip.cpp:507: undefined reference to `deflate'
+  $src/jdk/src/share/native/com/sun/java/util/jar/pack/zip.cpp:514: undefined reference to `deflateEnd'
+  $src/jdk/src/share/native/com/sun/java/util/jar/pack/zip.cpp:502: undefined reference to `deflate'
+  $src/jdk/src/share/native/com/sun/java/util/jar/pack/zip.cpp:518: undefined reference to `deflateEnd'
+  $build/jdk/objs/unpackexe/zip.o: In function `jar::get_crc32(unsigned int, unsigned char*, unsigned int)':
+  $src/jdk/src/share/native/com/sun/java/util/jar/pack/zip.cpp:61: undefined reference to `crc32'
+  $src/jdk/src/share/native/com/sun/java/util/jar/pack/zip.cpp:61: undefined reference to `crc32'
+  $src/jdk/src/share/native/com/sun/java/util/jar/pack/zip.cpp:61: undefined reference to `crc32'
+  $build/jdk/objs/unpackexe/zip.o: In function `gunzip::free()':
+  $src/jdk/src/share/native/com/sun/java/util/jar/pack/zip.cpp:645: undefined reference to `inflateEnd'
+  $build/jdk/objs/unpackexe/zip.o: In function `read_input_via_gzip(unpacker*, void*, long, long)':
+  $src/jdk/src/share/native/com/sun/java/util/jar/pack/zip.cpp:548: undefined reference to `inflate'
+  $src/jdk/src/share/native/com/sun/java/util/jar/pack/zip.cpp:548: undefined reference to `inflate'
+  $build/jdk/objs/unpackexe/zip.o: In function `gunzip::start(int)':
+  $src/jdk/src/share/native/com/sun/java/util/jar/pack/zip.cpp:637: undefined reference to `inflateInit2_'
+  collect2: error: ld returned 1 exit status
+
+(as happened on a Ubuntu 16.04 xenial host:
+  gcc --version
+  gcc (Ubuntu 5.4.0-6ubuntu1~16.04.9) 5.4.0 20160609
+  Copyright (C) 2015 Free Software Foundation, Inc.
+  This is free software; see the source for copying conditions.  There is NO
+  warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
+)
+
+The reason is that the order of command line arguments
+becomes important with --as-needed and the JDK build system
+places -lz at the beginning of the command line where it
+would normally place the zlib object files from its own
+bundled compiled version.
+
+Having -lz early in the command line is not useful, as it
+is discarded by the linker at that point in time as it
+hasn't seen any reference to the symbols provided yet.
+
+As it seems a generic pattern in the makefiles here, just
+place the $EXPECTED_OBJS early in the command line, before
+any additional libraries, so as to fix this once and for
+all.
+
+Signed-off-by: André Draszik <andre.draszik@jci.com>
+---
+Upstream-Status: Pending
+ make/common/NativeCompilation.gmk | 5 ++---
+ 1 file changed, 2 insertions(+), 3 deletions(-)
+
+diff --git a/make/common/NativeCompilation.gmk b/make/common/NativeCompilation.gmk
+index e7ea22ce..79f081fd 100644
+--- a/make/common/NativeCompilation.gmk
++++ b/make/common/NativeCompilation.gmk
+@@ -477,9 +477,8 @@ define SetupNativeCompilation
+ 
+     $$($1_TARGET) : $$($1_EXPECTED_OBJS) $$($1_RES) $$($1_GEN_MANIFEST)
+ 	$$(call LINKING_EXE_MSG,$$($1_BASENAME))
+-	$$($1_LDEXE) $$($1_LDFLAGS) $$($1_EXTRA_LDFLAGS) $(EXE_OUT_OPTION)$$($1_TARGET) \
+-	$$($1_EXPECTED_OBJS) $$($1_RES) $$($1_LDFLAGS_SUFFIX) \
+-	$$($1_EXTRA_LDFLAGS_SUFFIX)
++	$$($1_LDEXE) $$($1_EXPECTED_OBJS) $$($1_LDFLAGS) $$($1_EXTRA_LDFLAGS) $(EXE_OUT_OPTION)$$($1_TARGET) \
++	$$($1_RES) $$($1_LDFLAGS_SUFFIX) $$($1_EXTRA_LDFLAGS_SUFFIX)
+         ifneq (,$$($1_GEN_MANIFEST))
+ 	  $(MT) -nologo -manifest $$($1_GEN_MANIFEST) -outputresource:$$@;#1
+         endif
+-- 
+2.16.2
+
-- 
2.16.2

Re: [meta-java][PATCH 1/2] openjdk-8: fix MAKE detection patch

[Maxin B. John]

Hi André,

On Mon, Mar 12, 2018 at 04:38:05PM +0000, André Draszik wrote:
> From: André Draszik <andre.draszik@jci.com>
> 
> The patch had a few typos, leading to errors during ./configure
>    ../jdk8u-4be07cb28b21/common/autoconf/configure: line 8408: test: too many arguments
> 
> Change-Id: I867eba7aae3390aa869e69c86f29e77b505043e7
> ---
>  recipes-core/openjdk/patches-openjdk-8/dont-expect-fqpn-for-make.patch | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/recipes-core/openjdk/patches-openjdk-8/dont-expect-fqpn-for-make.patch b/recipes-core/openjdk/patches-openjdk-8/dont-expect-fqpn-for-make.patch
> index 6454237..5192d1a 100644
> --- a/recipes-core/openjdk/patches-openjdk-8/dont-expect-fqpn-for-make.patch
> +++ b/recipes-core/openjdk/patches-openjdk-8/dont-expect-fqpn-for-make.patch
> @@ -6,7 +6,7 @@
>       # User has supplied a make, test it.
>  -    if test ! -f "$MAKE"; then
>  -      AC_MSG_ERROR([The specified make (by MAKE=$MAKE) is not found.])
> -+    if test -a `dirname "$MAKE"` = "." -a ! -f "$MAKE"; then
> ++    if test `dirname "$MAKE"` = "." && ! test -f "$MAKE"; then
>  +      AC_PATH_PROGS(CHECK_MAKE, $MAKE)
>  +    else
>  +      CHECK_MAKE="${MAKE}"
> --

Merged, Thanks.

Best Regards,
Maxin

Re: [meta-java][PATCH 1/2] openjdk-8: fix MAKE detection patch

[André Draszik]

On Tue, 2018-03-13 at 18:10 +0200, Maxin B. John wrote:
> Hi André,
> 
> On Mon, Mar 12, 2018 at 04:38:05PM +0000, André Draszik wrote:
> > From: André Draszik <andre.draszik@jci.com>
> > 
> > The patch had a few typos, leading to errors during ./configure
> >    ../jdk8u-4be07cb28b21/common/autoconf/configure: line 8408: test: too
> > many arguments
> > 
> > Change-Id: I867eba7aae3390aa869e69c86f29e77b505043e7
> > ---
> >  recipes-core/openjdk/patches-openjdk-8/dont-expect-fqpn-for-make.patch
> > | 2 +-
> >  1 file changed, 1 insertion(+), 1 deletion(-)
> > 
> > diff --git a/recipes-core/openjdk/patches-openjdk-8/dont-expect-fqpn-
> > for-make.patch b/recipes-core/openjdk/patches-openjdk-8/dont-expect-
> > fqpn-for-make.patch
> > index 6454237..5192d1a 100644
> > --- a/recipes-core/openjdk/patches-openjdk-8/dont-expect-fqpn-for-
> > make.patch
> > +++ b/recipes-core/openjdk/patches-openjdk-8/dont-expect-fqpn-for-
> > make.patch
> > @@ -6,7 +6,7 @@
> >       # User has supplied a make, test it.
> >  -    if test ! -f "$MAKE"; then
> >  -      AC_MSG_ERROR([The specified make (by MAKE=$MAKE) is not found.])
> > -+    if test -a `dirname "$MAKE"` = "." -a ! -f "$MAKE"; then
> > ++    if test `dirname "$MAKE"` = "." && ! test -f "$MAKE"; then
> >  +      AC_PATH_PROGS(CHECK_MAKE, $MAKE)
> >  +    else
> >  +      CHECK_MAKE="${MAKE}"
> > --
> 
> Merged, Thanks.

Thanks Maxin.

Did you forget the 2nd patch, or is there an issue with it?

BTW, I just noticed that master and master-next have diverged quite a lot -
what is the plan wrt that?

Cheers,
Andre'

Re: [meta-java][PATCH 1/2] openjdk-8: fix MAKE detection patch

[Maxin B. John]

Hi André,

On Thu, Mar 15, 2018 at 09:10:05AM +0000, André Draszik wrote:
> On Tue, 2018-03-13 at 18:10 +0200, Maxin B. John wrote:
> > Hi André,
> > 
> > On Mon, Mar 12, 2018 at 04:38:05PM +0000, André Draszik wrote:
> > > From: André Draszik <andre.draszik@jci.com>
> > > 
> > > The patch had a few typos, leading to errors during ./configure
> > >    ../jdk8u-4be07cb28b21/common/autoconf/configure: line 8408: test: too
> > > many arguments
> > > 
> > > Change-Id: I867eba7aae3390aa869e69c86f29e77b505043e7
> > > ---
> > >  recipes-core/openjdk/patches-openjdk-8/dont-expect-fqpn-for-make.patch
> > > | 2 +-
> > >  1 file changed, 1 insertion(+), 1 deletion(-)
> > > 
> > > diff --git a/recipes-core/openjdk/patches-openjdk-8/dont-expect-fqpn-
> > > for-make.patch b/recipes-core/openjdk/patches-openjdk-8/dont-expect-
> > > fqpn-for-make.patch
> > > index 6454237..5192d1a 100644
> > > --- a/recipes-core/openjdk/patches-openjdk-8/dont-expect-fqpn-for-
> > > make.patch
> > > +++ b/recipes-core/openjdk/patches-openjdk-8/dont-expect-fqpn-for-
> > > make.patch
> > > @@ -6,7 +6,7 @@
> > >       # User has supplied a make, test it.
> > >  -    if test ! -f "$MAKE"; then
> > >  -      AC_MSG_ERROR([The specified make (by MAKE=$MAKE) is not found.])
> > > -+    if test -a `dirname "$MAKE"` = "." -a ! -f "$MAKE"; then
> > > ++    if test `dirname "$MAKE"` = "." && ! test -f "$MAKE"; then
> > >  +      AC_PATH_PROGS(CHECK_MAKE, $MAKE)
> > >  +    else
> > >  +      CHECK_MAKE="${MAKE}"
> > > --
> > 
> > Merged, Thanks.
> 
> Thanks Maxin.
> 
> Did you forget the 2nd patch, or is there an issue with it?

Ah, I missed it (or rather due to some weird reasons, it didn't
show up in my inbox)

> BTW, I just noticed that master and master-next have diverged quite a lot -
> what is the plan wrt that?

master-next was meant to be a place for testing the new patches. Will
update it soon.

> Cheers,
> Andre'

Cheers,
Maxin

Re: [meta-java][PATCH 1/2] openjdk-8: fix MAKE detection patch

[Maxin B. John]

Hi Andre',

On Thu, Mar 15, 2018 at 09:10:05AM +0000, André Draszik wrote:
> On Tue, 2018-03-13 at 18:10 +0200, Maxin B. John wrote:
> > Hi André,

<snip>

> > 
> > Merged, Thanks.
> 
> Thanks Maxin.
> 
> Did you forget the 2nd patch, or is there an issue with it?
> 
> BTW, I just noticed that master and master-next have diverged quite a lot -
> what is the plan wrt that?

Merged the 2nd patch and updated the master-next branch.

> Cheers,
> Andre'

Cheers,
Maxin