Upgrade to Classpath 0.99

Fixed this quickly to avoid problem of work being done on old Classpath version. Patch in r15409.

classpath 0.98 build is broken on aix. As a result, all nightly regression tests on piccolo are failing.

This may be quite annoying to debug as configuring classpath on aix takes well over an hour.

build:
[available] Found: /usr/gnu/bin/make
Property "config.portable-native-sync" has not been set
[available] Unable to find /home/jikesrvm/components/classpath/98p0/classpath/Ma
kefile
[property] Loading Environment env.
[exec] Current OS is AIX
[exec] Setting environment variable: CXX=/usr/gnu/bin/g++ -w -O -pthread -W
a,-mppc
[exec] Setting environment variable: CC=/usr/gnu/bin/gcc -w -O -pthread -Wa
,-mppc
[exec] Setting environment variable: OBJECT_MODE=32
[exec] Setting environment variable: JAVAC=/home/jikesrvm/components/ecj/3.
2/ecj-3.2/ecj
[exec] Executing '/home/jikesrvm/components/classpath/98p0/classpath/config
ure' with arguments:
[exec] '--enable-jni'
[exec] '--disable-plugin'
[exec] '--disable-gconf-peer'
[exec] '--disable-examples'
[exec] '--disable-gtk-peer'
[exec] 'CPPFLAGS=-D_POSIX_SOURCE -D_ALL_SOURCE -D_LINUX_SOURCE_COMPAT'
[exec]
[exec] The ' characters around the executable and arguments are
[exec] not part of the command.
[exec] checking build system type... powerpc-ibm-aix5.3.0.0
[exec] checking host system type... powerpc-ibm-aix5.3.0.0
[exec] checking target system type... powerpc-ibm-aix5.3.0.0
[exec] checking for a BSD-compatible install... ./install-sh -c
[exec] checking whether build environment is sane... yes
[exec] checking for a thread-safe mkdir -p... ./install-sh -c -d
[exec] checking for gawk... gawk
[exec] checking whether make sets $(MAKE)... yes
[exec] checking how to create a ustar tar archive... gnutar
[exec] checking for style of include used by make... GNU
[exec] checking for gcc... /usr/gnu/bin/gcc -w -O -pthread -Wa,-mppc
[exec] checking for C compiler default output file name... a.out
[exec] checking whether the C compiler works... yes
[exec] checking whether we are cross compiling... no
[exec] checking for suffix of executables...
[exec] checking for suffix of object files... o
[exec] checking whether we are using the GNU C compiler... yes
[exec] checking whether /usr/gnu/bin/gcc -w -O -pthread -Wa,-mppc accepts -
g... yes
[exec] checking for /usr/gnu/bin/gcc -w -O -pthread -Wa,-mppc option to acc
ept ISO C89... none needed
[exec] checking dependency style of /usr/gnu/bin/gcc -w -O -pthread -Wa,-mp
pc... gcc3
[exec] checking how to run the C preprocessor... /usr/gnu/bin/gcc -w -O -pt
hread -Wa,-mppc -E
[exec] checking for grep that handles long lines and -e... /usr/gnu/bin/gre
p
[exec] checking for egrep... /usr/gnu/bin/grep -E
[exec] checking for ANSI C header files... yes
[exec] checking for sys/types.h... yes
[exec] checking for sys/stat.h... yes
[exec] checking for stdlib.h... yes
[exec] checking for string.h... yes
[exec] checking for memory.h... yes
[exec] checking for strings.h... yes
[exec] checking for inttypes.h... yes
[exec] checking for stdint.h... yes
[exec] checking for unistd.h... yes
[exec] checking alsa/asoundlib.h usability... no
[exec] checking alsa/asoundlib.h presence... no
[exec] checking for alsa/asoundlib.h... no
[exec] checking dssi.h usability... no
[exec] checking dssi.h presence... no
[exec] checking for dssi.h... no
[exec] checking whether to regenerate the headers... no
[exec] checking whether to regenerate the GJDoc parser... yes
[exec] checking whether ln -s works... yes
[exec] checking for a BSD-compatible install... ./install-sh -c
[exec] checking whether we are using the GNU C++ compiler... yes
[exec] checking whether /usr/gnu/bin/g++ -w -O -pthread -Wa,-mppc accepts -
g... yes
[exec] checking dependency style of /usr/gnu/bin/g++ -w -O -pthread -Wa,-mp
pc... gcc3
[exec] checking for a sed that does not truncate output... /usr/gnu/bin/sed
[exec] checking for ld used by /usr/gnu/bin/gcc -w -O -pthread -Wa,-mppc...
/usr/bin/ld
[exec] checking if the linker (/usr/bin/ld) is GNU ld... no
[exec] checking for /usr/bin/ld option to reload object files... -r
[exec] checking for BSD-compatible nm... /usr/bin/nm -B
[exec] checking how to recognize dependent libraries... pass_all
[exec] checking dlfcn.h usability... yes
[exec] checking dlfcn.h presence... yes
[exec] checking for dlfcn.h... yes
[exec] checking how to run the C++ preprocessor... /usr/gnu/bin/g++ -w -O -
pthread -Wa,-mppc -E
[exec] checking for g77... no
[exec] checking for xlf... no
[exec] checking for f77... no
[exec] checking for frt... no
[exec] checking for pgf77... no
[exec] checking for cf77... no
[exec] checking for fort77... no
[exec] checking for fl32... no
[exec] checking for af77... no
[exec] checking for xlf90... no
[exec] checking for f90... no
[exec] checking for pgf90... no
[exec] checking for pghpf... no
[exec] checking for epcf90... no
[exec] checking for gfortran... gfortran
[exec] checking whether we are using the GNU Fortran 77 compiler... yes
[exec] checking whether gfortran accepts -g... yes
[exec] checking the maximum length of command line arguments... 18432
[exec] checking command to parse /usr/bin/nm -B output from /usr/gnu/bin/gc
c -w -O -pthread -Wa,-mppc object... ok
[exec] checking for objdir... .libs
[exec] checking for ar... ar
[exec] checking for ranlib... ranlib
[exec] checking for strip... strip
[exec] checking for correct ltmain.sh version... yes
[exec] checking if /usr/gnu/bin/gcc w -O -pthread -Wa,-mppc supports -fno
rtti -fno-exceptions... yes
[exec] checking for /usr/gnu/bin/gcc -w -O -pthread -Wa,-mppc option to pro
duce PIC...
[exec] checking if /usr/gnu/bin/gcc -w -O -pthread -Wa,-mppc static flag -s
tatic works... no
[exec] checking if /usr/gnu/bin/gcc -w -O -pthread -Wa,-mppc supports -c -o
file.o... yes
[exec] checking whether the /usr/gnu/bin/gcc -w -O -pthread -Wa,-mppc linke
r (/usr/bin/ld) supports shared libraries... yes
[exec] checking dynamic linker characteristics... aix5.3.0.0 ld.so
[exec] checking how to hardcode library paths into programs... immediate
[exec] checking whether stripping libraries is possible... no
[exec] checking if libtool supports shared libraries... yes
[exec] checking whether to build shared libraries... yes
[exec] checking whether to build static libraries... no
[exec] configure: creating libtool
[exec] appending configuration tag "CXX" to libtool
[exec] checking for ld used by /usr/gnu/bin/g++ -w -O -pthread -Wa,-mppc...
/usr/bin/ld
[exec] checking if the linker (/usr/bin/ld) is GNU ld... no
[exec] checking whether the /usr/gnu/bin/g++ -w -O -pthread -Wa,-mppc linke
r (/usr/bin/ld) supports shared libraries... yes
[exec] checking for /usr/gnu/bin/g++ -w -O -pthread -Wa,-mppc option to pro
duce PIC...
[exec] checking if /usr/gnu/bin/g++ -w -O -pthread -Wa,-mppc static flag -s
tatic works... no
[exec] checking if /usr/gnu/bin/g++ -w -O -pthread -Wa,-mppc supports -c -o
file.o... yes
[exec] checking whether the /usr/gnu/bin/g++ -w -O -pthread -Wa,-mppc linke
r (/usr/bin/ld) supports shared libraries... yes
[exec] checking dynamic linker characteristics... aix5.3.0.0 ld.so
[exec] (cached) (cached) checking how to hardcode library paths into progra
ms... immediate
[exec] appending configuration tag "F77" to libtool
[exec] checking if libtool supports shared libraries... yes
[exec] checking whether to build shared libraries... yes
[exec] checking whether to build static libraries... no
[exec] checking for gfortran option to produce PIC...
[exec] checking if gfortran static flag -static works... no
[exec] checking if gfortran supports -c -o file.o... yes
[exec] checking whether the gfortran linker (/usr/bin/ld) supports shared l
ibraries... yes
[exec] checking dynamic linker characteristics... aix5.3.0.0 ld.so
[exec] (cached) (cached) checking how to hardcode library paths into progra
ms... immediate
[exec] checking for gawk... (cached) gawk
[exec] checking for gcc... (cached) /usr/gnu/bin/gcc -w -O -pthread -Wa,-mp
pc
[exec] checking whether we are using the GNU C compiler... (cached) yes
[exec] checking whether /usr/gnu/bin/gcc -w -O -pthread -Wa,-mppc accepts -
g... (cached) yes
[exec] checking for /usr/gnu/bin/gcc -w -O -pthread -Wa,-mppc option to acc
ept ISO C89... (cached) none needed
[exec] checking dependency style of /usr/gnu/bin/gcc -w -O -pthread -Wa,-mp
pc... (cached) gcc3
[exec] checking whether /usr/gnu/bin/gcc -w -O -pthread -Wa,-mppc and cc un
derstand -c and -o together... yes
[exec] checking how to run the C preprocessor... /usr/gnu/bin/gcc -w -O -pt
hread -Wa,-mppc -E
[exec] checking _attribute_((,,))... yes
[exec] checking _attribute_((unused))... yes
[exec] checking for ANSI C header files... (cached) yes
[exec] checking size of void *... 4
[exec] checking whether byte ordering is bigendian... yes
[exec] checking for unistd.h... (cached) yes
[exec] checking for sys/types.h... (cached) yes
[exec] checking sys/config.h usability... no
[exec] checking sys/config.h presence... no
[exec] checking for sys/config.h... no
[exec] checking sys/ioctl.h usability... yes
[exec] checking sys/ioctl.h presence... yes
[exec] checking for sys/ioctl.h... yes
[exec] checking asm/ioctls.h usability... no
[exec] checking asm/ioctls.h presence... no
[exec] checking for asm/ioctls.h... no
[exec] checking for inttypes.h... (cached) yes
[exec] checking for stdint.h... (cached) yes
[exec] checking utime.h usability... yes
[exec] checking utime.h presence... yes
[exec] checking for utime.h... yes
[exec] checking sys/utime.h usability... no
[exec] checking sys/utime.h presence... no
[exec] checking for sys/utime.h... no
[exec] checking sys/filio.h usability... no
[exec] checking sys/filio.h presence... no
[exec] checking for sys/filio.h... no
[exec] checking sys/time.h usability... yes
[exec] checking sys/time.h presence... yes
[exec] checking for sys/time.h... yes
[exec] checking sys/select.h usability... yes
[exec] checking sys/select.h presence... yes
[exec] checking for sys/select.h... yes
[exec] checking crt_externs.h usability... no
[exec] checking crt_externs.h presence... no
[exec] checking for crt_externs.h... no
[exec] checking fcntl.h usability... yes
[exec] checking fcntl.h presence... yes
[exec] checking for fcntl.h... yes
[exec] checking sys/mman.h usability... yes
[exec] checking sys/mman.h presence... yes
[exec] checking for sys/mman.h... yes
[exec] checking magic.h usability... no
[exec] checking magic.h presence... no
[exec] checking for magic.h... no
[exec] checking sys/event.h usability... no
[exec] checking sys/event.h presence... no
[exec] checking for sys/event.h... no
[exec] checking sys/epoll.h usability... no
[exec] checking sys/epoll.h presence... no
[exec] checking for sys/epoll.h... no
[exec] checking ifaddrs.h usability... no
[exec] checking ifaddrs.h presence... no
[exec] checking for ifaddrs.h... no
[exec] checking netinet/in_systm.h usability... yes
[exec] checking netinet/in_systm.h presence... yes
[exec] checking for netinet/in_systm.h... yes
[exec] checking netinet/ip.h usability... yes
[exec] checking netinet/ip.h presence... yes
[exec] checking for netinet/ip.h... yes
[exec] checking net/if.h usability... yes
[exec] checking net/if.h presence... yes
[exec] checking for net/if.h... yes
[exec] checking sys/loadavg.h usability... no
[exec] checking sys/loadavg.h presence... no
[exec] checking for sys/loadavg.h... no
[exec] checking sys/sockio.h usability... no
[exec] checking sys/sockio.h presence... no
[exec] checking for sys/sockio.h... no
[exec] checking for library containing inet_pton... none required
[exec] checking for gethostname in -lsocket... no
[exec] checking for ftruncate... yes
[exec] checking for fsync... yes
[exec] checking for select... yes
[exec] checking for gethostname... yes
[exec] checking for socket... yes
[exec] checking for strerror... yes
[exec] checking for fork... yes
[exec] checking for pipe... yes
[exec] checking for execve... yes
[exec] checking for open... yes
[exec] checking for close... yes
[exec] checking for lseek... yes
[exec] checking for fstat... yes
[exec] checking for read... yes
[exec] checking for readv... yes
[exec] checking for write... yes
[exec] checking for writev... yes
[exec] checking for htonl... yes
[exec] checking for memset... yes
[exec] checking for htons... yes
[exec] checking for connect... yes
[exec] checking for getsockname... yes
[exec] checking for getpeername... yes
[exec] checking for bind... yes
[exec] checking for listen... yes
[exec] checking for accept... yes
[exec] checking for recvfrom... yes
[exec] checking for send... yes
[exec] checking for sendto... yes
[exec] checking for setsockopt... yes
[exec] checking for getsockopt... yes
[exec] checking for time... yes
[exec] checking for mktime... yes
[exec] checking for gethostbyname_r... yes
[exec] checking for localtime_r... yes
[exec] checking for strerror_r... yes
[exec] checking for fcntl... yes
[exec] checking for statvfs... yes
[exec] checking for mmap... yes
[exec] checking for munmap... yes
[exec] checking for mincore... yes
[exec] checking for msync... yes
[exec] checking for madvise... yes
[exec] checking for getpagesize... yes
[exec] checking for sysconf... yes
[exec] checking for lstat... yes
[exec] checking for readlink... yes
[exec] checking for inet_aton... yes
[exec] checking for inet_addr... yes
[exec] checking for inet_pton... yes
[exec] checking for getifaddrs... no
[exec] checking for kqueue... no
[exec] checking for kevent... no
[exec] checking for epoll_create... no
[exec] checking for getloadavg... no
[exec] checking for magic_open in -lmagic... no
[exec] checking whether struct sockaddr_in6 is in netinet/in.h... yes
[exec] checking whether time.h and sys/time.h may both be included... yes
[exec] checking whether struct tm is in sys/time.h or time.h... time.h
[exec] checking for struct tm.tm_zone... no
[exec] checking whether tzname is declared... yes
[exec] checking for tzname... yes
[exec] checking for tm_gmtoff in struct tm... no
[exec] checking for global timezone variable... yes
[exec] checking for an ANSI C-conforming const... yes
[exec] checking for inline... inline
[exec] checking for _attribute_... yes
[exec] checking how many arguments gethostbyname_r() takes... six
[exec] checking for ld used by GCC... /usr/bin/ld
[exec] checking if the linker (/usr/bin/ld) is GNU ld... no
[exec] /bin/sh: ./config.rpath: not found
[exec] checking for shared library run path origin... done
[exec] checking for iconv... yes
[exec] checking how to link with libiconv... -liconv
[exec] checking for iconv declaration...
[exec] extern size_t iconv (iconv_t cd, char * *inbuf, size_t *inb
ytesleft, char * *outbuf, size_t *outbytesleft);
[exec] checking for MSG_NOSIGNAL... no
[exec] checking for SO_NOSIGPIPE ... no
[exec] checking for MSG_WAITALL... yes
[exec] checking for __gmpz_mul_si in -lgmp... no
[exec] checking gmp.h usability... yes
[exec] checking gmp.h presence... yes
[exec] checking for gmp.h... yes
[exec] configure: WARNING: no
[exec] checking jni_md.h support... checking whether to enable maintainer-s
pecific portions of Makefiles... no
[exec] checking for mkdir... /usr/bin/mkdir
[exec] checking for cp... /usr/bin/cp
[exec] checking for date... /usr/bin/date
[exec] checking for find... /usr/gnu/bin/find
[exec] checking for zip... no
[exec] checking for a jar-like tool... trying fastjar, gjar and jar
[exec] checking for fastjar... no
[exec] checking for gjar... no
[exec] checking for jar... /usr/java5/bin/jar
[exec] checking whether to regenerate parsers with jay... no
[exec] checking for stdint types... stdint.h (shortcircuit)
[exec] head: `-1' option is obsolete; use `-n 1'
[exec] Try `head --help' for more information.
[exec] make use of stdint.h in include/config-int.h (assuming C99 compatibl
e system)
[exec] checking for cacao... no
[exec] checking for jamvm... no
[exec] checking for kaffe... no
[exec] checking for gij... no
[exec] checking for java... java
[exec] checking for uudecode... yes
[exec] checking if uudecode can decode base 64 file... yes
[exec] checking if java works... yes
[exec] /home/jikesrvm/components/classpath/98p0/classpath/configure[35072]:
test: argument expected
[exec] configure: error: specified ANTLR jar file not found.
[exec] checking for the ANTLR parser generator JAR file...
[antcall] Exiting /sanity/ppc32-aix/jikesrvm/core/jikesrvm/build/components/cl

Hopeful fix in r15413.

Still broken:

[exec] make[3]: Entering directory `/home/jikesrvm/components/classpath/98p0/classpath/native/jni/classpath'
[exec] /bin/sh ../../../libtool --tag=CC --mode=compile /usr/gnu/bin/gcc -w -O -pthread -Wa,-mppc -DHAVE_CONFIG_H -I. -I../../../include -I../../../include -I../../../native/jni/classpath -I../../../native/jni/native-lib -D_POSIX_SOURCE -D_ALL_SOURCE -D_LINUX_SOURCE_COMPAT -W -Wall -Wmissing-declarations -Wwrite-strings -Wmissing-prototypes -Wno-long-long -Wstrict-prototypes -pedantic -fexceptions -fasynchronous-unwind-tables -g -O2 -MT jcl.lo -MD -MP -MF .deps/jcl.Tpo -c -o jcl.lo jcl.c
[exec] mkdir .libs
[exec] /usr/gnu/bin/gcc -w -O -pthread -Wa,-mppc -DHAVE_CONFIG_H -I. -I../../../include -I../../../include -I../../../native/jni/classpath -I../../../native/jni/native-lib -D_POSIX_SOURCE -D_ALL_SOURCE -D_LINUX_SOURCE_COMPAT -W -Wall -Wmissing-declarations -Wwrite-strings -Wmissing-prototypes -Wno-long-long -Wstrict-prototypes -pedantic -fexceptions -fasynchronous-unwind-tables -g -O2 -MT jcl.lo -MD -MP -MF .deps/jcl.Tpo -c jcl.c -DPIC -o .libs/jcl.o
[exec] 5668 problems (5668 warnings)In file included from jcl.c:43:
[exec] ./jcl.h:50:2: error: #error "Unknown pointer size"
[exec] jcl.c:94:2: error: #error "Pointer size is not supported."
[exec] make[3]: *** [jcl.lo] Error 1

http://jikesrvm.anu.edu.au/cattrack/results/piccolo.watson.ibm.com/core-ppc32/7729/prototype/Output.txt

The error relates to SIZEOF_VOID_P not being set. The relevant bit of configure is:

int
main ()
{
switch (0) case 0: case (sizeof (void *) == $ac_size):;
;
return 0;
}

where ac_size is a guess at the size of the pointer. It is likely that AIX is choosing a value that isn't 4 or 8 here. Dave, could you look at components/classpath/98p0/classpath/include/config.h to see what size of pointer configure is creating? The relevant bits of configure don't appear to have changed from 0.97 to 0.98 so it is odd AIX is now breaking.

the configure step is failing:
.......
[exec] checking how to run the C preprocessor... /usr/gnu/bin/gcc -w -O -pthread -Wa,-mppc -E
[exec] grep: conftest.er1: No such file or directory
[exec] checking _attribute_((,,))... no
[exec] checking _attribute_((unused))... no
[exec] checking for ANSI C header files... (cached) yes
[exec] cat: conftest.err: No such file or directory
[exec] sed: can't read conftest.c: No such file or directory
[exec] sed: can't read conftest.c: No such file or directory
[exec] configure: error: cannot determine a size for void *
[exec] checking size of void *...

therefore config.h is not being created.

There's a slight chance this is happening because /tmp is being filled up by lots of temporary files created by the configure script and therefore it can't create conftest.c. I'm going to up the partition size allocated for /tmp to see if that helps...long shot...

bulk defer open issues to 3.1.2

What's the current status of this?

GNU Classpath 0.99 was released in March 2012. We need to evaluate whether it makes sense switching to 0.99 instead of 0.98.

I'm going to change the issue title to reflect 0.99 being released, but since good progress has been made on switching to the OpenJDK libraries I don't think it is worth pursuing the upgrade to Classpath 0.99.

Attached is a patch (against revision 8aaadb754ec2) that performs the upgrade to Classpath 0.99. Issue RVM-781 is still present, though.

Nevertheless, the patch should be helpful in tracking any improvements to Classpath made upstream, as it adds a classpath.from-git property (Classpath no longer uses CVS) which, when true, fetches the latest development version of classpath.

Hi Andreas, Could you separate out the part of the patch that makes the change to get classpath from git instead of cvs? That would be useful to apply independent of the rest of this. Thanks!

The attached classpath-from-git.patch contains only the changes necessary to obtain the Classpath head from Git rather than CVS (by setting -Dclasspath.from-git=true). It does not change the classpath.from-web behaviour in any way. Also, it does not yet bump the version number from 0.97.2 to 0.99, as the classpath.from-web case still uses 0.97.2. (This is in line with current behaviour, but still slightly wrong; IMHO, the classpath.from-git/cvs case should place its results in components/classpath/head rather than components/classpath/${classpath.version}p${classpath.patchlevel}/.)

Applied classpath-from-git-patch in 10457:4cf370c13a87. Thanks Andreas!

(1) we should go ahead and hg rm obsolete patches
(2) Makes sense. components/classpath-git or something like that?
(3) Yep. Once RVM-781 gets solved it makes sense to move to the newer version.

OK, I've addressed issues (1) and (2); obsolete patches have been removed and the head is now kept in components/classpath/HEADp${classpath.patchlevel}.

You can review the changes I made on Bitbucket: https://bitbucket.org/sewe/jikesrvm/changesets/tip/..bookmark%28%22rvm-773%22%29. To retrieve and merge the changes, simply pull from https://bitbucket.org/sewe/jikesrvm.

I pulled changeset 7f209dcb71a0 into the main repository. The other two changesets (c5ec6e31d8e2 and 47362bbbdcf8) caused problems for me. "c5" caused the patches to not all apply cleanly (some of them are built on top of others I think) and "47" caused classpath to be rebuilt from scratch everytime I built jikesrvm (something went wrong with the check to see if classpath was already built). I didn't look into either failure very deeply, so they might be not that hard to fix.

Hm, c5ec6e31d8e2 not applying the patches cleanly is odd; it works on my machine (both with -Dclasspath.from-web=true and -Dclasspath.from-git=true, albeit in a few cases only with adjusting offsets and/or fuzz:

    [patch] patching file java/lang/reflect/Constructor.java
    [patch] patching file java/lang/reflect/Field.java
    [patch] Hunk #9 succeeded at 416 (offset 1 line).
    [patch] patching file java/lang/reflect/Method.java
    [patch] patching file vm/reference/java/lang/reflect/VMConstructor.java
    [patch] patching file vm/reference/java/lang/reflect/VMField.java
    [patch] patching file vm/reference/java/lang/reflect/VMMethod.java
    [patch] patching file java/util/ArrayList.java
    [patch] patching file java/util/Vector.java
    [patch] Hunk #1 succeeded at 903 (offset -6 lines).
    [patch] Hunk #2 succeeded at 918 (offset -6 lines).
    [patch] patching file gnu/xml/transform/ParameterNode.java
    [patch] patching file gnu/xml/transform/Stylesheet.java
    [patch] patching file gnu/xml/transform/Template.java
    [patch] patching file gnu/xml/transform/TemplatesImpl.java
    [patch] patching file gnu/xml/transform/CallTemplateNode.java
    [patch] patching file gnu/xml/transform/Stylesheet.java
    [patch] Hunk #1 succeeded at 1690 (offset 10 lines).
    [patch] patching file gnu/xml/transform/ForEachNode.java
    [patch] patching file gnu/xml/transform/SortKey.java
    [patch] patching file gnu/xml/transform/Stylesheet.java
    [patch] Hunk #1 succeeded at 1654 with fuzz 1 (offset 10 lines).
    [patch] patching file gnu/xml/transform/TemplateNode.java
    [patch] patching file gnu/xml/dom/DomAttr.java
    [patch] patching file gnu/xml/dom/DomDocument.java
    [patch] patching file gnu/xml/dom/DomElement.java
    [patch] patching file gnu/xml/dom/DomNsNode.java
    [patch] patching file gnu/xml/dom/ls/DomLSParser.java
    [patch] patching file gnu/xml/dom/ls/SAXEventSink.java
    [patch] patching file gnu/xml/transform/XSLURIResolver.java

Maybe your patch (using GNU patch 2.6 here) is more strict about these things.

Please test with changeset ad12193c405e at https://bitbucket.org/sewe/jikesrvm/changeset/ad12193c405e; I tweaked the patches such that the apply without fuzz now.

bulk defer to 3.1.4