On 2015-10-15 15:02, Prof Brian Ripley wrote:> On 15/10/2015 13:32, Michael Felt wrote:
>> Hi.
>>
>> Just wanted to let you know I am getting close to packaging R for AIX
in
>> 64.bit mode.
>
> Which version? (You mentioned 3.1.3 and 3.2.2 far below.) There is
> little value in reporting on frozen branches, and most value in
> reporting on R-devel where all the current changes have been
> incorporated.
This report is based on R-3.2.2 I shall repeat asap with R-devel and
report any significant differences.
compiler is:
root at x072:[/data/prj/cran/64/R-3.2.2/src/main]gcc --version
gcc (GCC) 4.7.4
Copyright (C) 2012 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.
root at x072:[/data/prj/cran/64/R-3.2.2/src/main]gfortran --version
GNU Fortran (GCC) 4.7.4
Copyright (C) 2012 Free Software Foundation, Inc.
AIX level:
root at x072:[/data/prj/cran/64/R-3.2.2/src/main]oslevel -s
7100-03-05-1524
Non-IBM filesets installed:
root at x072:[/data/prj/cran/64/R-3.2.2/src/main]lslpp -L | grep aixtools
aixtools.buildaix.man 2.0.1.0 C F man pages
aixtools.buildaix.man.en_US
aixtools.buildaix.rte 2.0.3.1 C F 1540 1007 2004
aixtools.gmp.rte 5.0.5.0 C F gmp version 5.0.5
for powerpc
aixtools.gmp.share 5.0.5.0 C F gmp share and
documentation
aixtools.gnu.bash.man.en_US
aixtools.gnu.bash.rte 4.3.42.0 C F 1537 0913 1505
aixtools.gnu.bash.share 4.3.42.0 C F universal pages
aixtools.gnu.coreutils.rte
aixtools.gnu.gettext.rte 0.19.5.0 C F 1531 0807 0000
aixtools.gnu.gettext.share
aixtools.gnu.make.rte 4.1.0.0 C F 1523 0613 1158
aixtools.gnu.wget.man.en_US
aixtools.gnu.wget.rte 1.16.3.0 C F 1530 0731 0651
aixtools.gnu.wget.share 1.16.3.0 C F universal pages
aixtools.libiconv.man 1.14.0.0 C F built 1447 1127 1728
aixtools.libiconv.rte 1.14.0.0 C F libiconv 1447 1127 1728
aixtools.libiconv.share 1.14.0.0 C F built 1447 1127 1728
aixtools.mpc.rte 1.0.2.0 C F mpc 2014.Oct.31
15.44 CUT
aixtools.mpc.share 1.0.2.0 C F share and documentation
aixtools.mpfr.rte 2.4.2.0 C F mpfr 2014.Oct.31
15.19 CUT
aixtools.mpfr.share 2.4.2.0 C F share and documentation
aixtools.obj.gcc.man 4.7.4.0 C F man pages
aixtools.obj.gcc.rte 4.7.4.0 C F 1515 0416 2134
aixtools.obj.gcc.share 4.7.4.0 C F universal pages
aixtools64.gnu.libiconv.rte
>> export OBJECT_MODE=64
>> export CFLAGS="-maix64 -O2"
>> export FFLAGS="-maix64 -O2"
>
> See the manual: flags such as -maix64 should be part of CC, not
> CFLAGS. That explains a lot of what was reported.
> Specifically,
>
https://cran.r-project.org/doc/manuals/r-patched/R-admin.html#Essential-programs-and-libraries
> and https://cran.r-project.org/doc/manuals/r-patched/R-admin.html#AIX .
>
> It is safer to put such things on the configure command line or in a
> config.site file (they will be found when updating if you do).
>
Not sure what you mean by a config.site file - perhaps what I use as a
"setup.env" script.
I followed the suggestion to move the flag -maix64 from CFLAGS and
FFLAGS (where I think it belongs) to the compiler defintions. It does
build better, but, still many concerns.
Details - too many lines of duplicate symbol references to paste (not
wishing to spam the list - for details see
http://download.aixtools.net/test/R/R-3.2.2-main-nohup.out> [Lots of test output removed.]
>
The summary is:
ld: 0711-318 ERROR: Undefined symbols were found.
The following symbols are in error:
Symbol Inpndx TY CL Source-File(Object-File) OR
Import-File{Shared-object}
RLD: Address Section Rld-type
Referencing Symbol
----------------------------------------------------------------------------------------------
.pthread_mutex_lock [24] ER PR
pcre_jit_compile.c(../extra/pcre/libpcre.a[pcre_jit_compile.o])
000076f8 .text R_RBR [118]
<.sljit_free_exec>
00007940 .text R_RBR [120]
<.sljit_generate_code>
00028c30 .text R_RBR [230]
.pcre_jit_free_unused_memory
.pthread_mutex_unlock [26] ER PR
pcre_jit_compile.c(../extra/pcre/libpcre.a[pcre_jit_compile.o])
0000777c .text R_RBR [118]
<.sljit_free_exec>
000079cc .text R_RBR [120]
<.sljit_generate_code>
00007b94 .text R_RBR [120]
<.sljit_generate_code>
00008038 .text R_RBR [120]
<.sljit_generate_code>
00028cc4 .text R_RBR [230]
.pcre_jit_free_unused_memory
ER: The return code is 8.
ld: 0711-415 WARNING: Symbol BM_norm_keep is already exported.
ld: 0711-415 WARNING: Symbol LoadInitFile is already exported.
ld: 0711-415 WARNING: Symbol N01_kind is already exported.
ld: 0711-415 WARNING: Symbol R_ReadItemDepth is already exported.
ld: 0711-224 WARNING: Duplicate symbol: BM_norm_keep
ld: 0711-224 WARNING: Duplicate symbol: N01_kind
ld: 0711-224 WARNING: Duplicate symbol: LoadInitFile
ld: 0711-224 WARNING: Duplicate symbol: R_ReadItemDepth
ld: 0711-317 ERROR: Undefined symbol: .pthread_mutex_lock
ld: 0711-317 ERROR: Undefined symbol: .pthread_mutex_unlock
collect2: error: ld returned 8 exit status
The undefined sysbols can be resolved by adding -lpthread to the gcc
command (and can be automated by adding that to LDFLAGS I expect, but
imho - better if that gets caught by configure).
It then ends with:
RESOLVE: 6793 of 18448 symbols were kept.
RESOLVE: The return code is 4.
(ld): addgl /usr/lib/glink64.o
ADDGL: Glink code added for 212 symbols.
(ld): er full
ER: There are no unresolved symbols.
(ld): mismatch
MISMATCH: No type mismatches exist.
(ld): comprld
COMPRLD: Kept 52596 of 52596 relocation entries.
(ld): origin page 0x100000000 0x110000000
(ld): libpath
/usr/local/lib:/opt/lib/gcc/powerpc-ibm-aix5.3.0.0/4.7.4/ppc64:/opt/lib/gcc/powerpc-ibm-aix5.3.0.0/4.7.4/../../../ppc64:/opt/lib/gcc/powerpc-ibm-aix5.3.0.0/4.7.4:/opt/lib/gcc/powerpc-ibm-aix5.3.0.0/4.7.4/../../..:/usr/lib:/lib
LIBPATH: Library path set to
/usr/local/lib:/opt/lib/gcc/powerpc-ibm-aix5.3.0.0/4.7.4/ppc64:/opt/lib/gcc/powerpc-ibm-aix5.3.0.0/4.7.4/../../../ppc64:/opt/lib/gcc/powerpc-ibm-aix5.3.0.0/4.7.4:/opt/lib/gcc/powerpc-ibm-aix5.3.0.0/4.7.4/../../..:/usr/lib:/lib
(ld): save 1L .
SAVE: Section sizes = 3374479+365673+669656 (0x337D8F+0x59469+0xA37D8 hex)
SAVE: Size of TOC: 5208 (0x1458 hex)
SAVE: No .loader section relocation entries exist for the .text section.
(ld): rc
RC: Highest return code was 4.
ld: 0711-415 WARNING: Symbol BM_norm_keep is already exported.
ld: 0711-415 WARNING: Symbol LoadInitFile is already exported.
ld: 0711-415 WARNING: Symbol N01_kind is already exported.
ld: 0711-415 WARNING: Symbol R_ReadItemDepth is already exported.
ld: 0711-224 WARNING: Duplicate symbol: BM_norm_keep
ld: 0711-224 WARNING: Duplicate symbol: N01_kind
ld: 0711-224 WARNING: Duplicate symbol: LoadInitFile
ld: 0711-224 WARNING: Duplicate symbol: R_ReadItemDepth
a) There is still a non-zero status (rc=4)
b) unhappy about all the duplications, even double exports