Disable ENABLE_PADLOCK_SUPPORT?
Rob Goedman
Brand new to Sage builds!
On Mac OS 10.5.2, with:
Robs-Intel:sage-3.0.rc0 rob$ gcc -v
Using built-in specs.
Target: i686-apple-darwin9
Configured with: /var/tmp/gcc_42/gcc_42-5531~1/src/configure --disable-
checking -enable-werror --prefix=/usr --mandir=/usr/share/man --enable-
languages=c,objc,c++,obj-c++ --program-transform-name=/^[cg][^.-]*$/s/
$/-4.2/ --with-gxx-include-dir=/usr/include/c++/4.0.0 --with-slibdir=/
usr/lib --build=i686-apple-darwin9 --host=i686-apple-darwin9 --
target=i686-apple-darwin9
Thread model: posix
gcc version 4.2.1 (Apple Inc. build 5531)
make gives below (summarized) error on 2.11, alpha6, rc0 and rc1.
Below attachment contains the complete rc1 install.log .
If, as a quick test, the Mac specific '-fasm-blocks' flag is added in
the subdir cipher, make complains about the assembler code in the asm
block in poll_padlock.
Do I have to disable ENABLE_PADLOCK_SUPPORT? If so, can I force the
sage make to use './configure -disable_padlock_support'?
As this is not related to the upcoming Sage 3.0 release, I'm fine to
wait for a binary release, although ultimately I would like to be able
to build sage myself.
Thanks,
Rob
----------------------------------------------
Making all in cipher
/bin/sh ../libtool --tag=CC --mode=compile gcc -DHAVE_CONFIG_H -I. -
I.. -I../src -I../src -I/Users/rob/Downloads/sage-3.0.rc0/local/
include -I/Users/rob/Downloads/sage-3.0.rc0/local/include -g -O2 -Wall
-Wpointer-arith -MT cipher.lo -MD -MP -MF .deps/cipher.Tpo -c -o
cipher.lo cipher.c
mkdir .libs
gcc -DHAVE_CONFIG_H -I. -I.. -I../src -I../src -I/Users/rob/
Downloads/sage-3.0.rc0/local/include -I/Users/rob/Downloads/
sage-3.0.rc0/local/include -g -O2 -Wall -Wpointer-arith -MT cipher.lo -
MD -MP -MF .deps/cipher.Tpo -c cipher.c -fno-common -DPIC -o .libs/
cipher.o
gcc -DHAVE_CONFIG_H -I. -I.. -I../src -I../src -I/Users/rob/
Downloads/sage-3.0.rc0/local/include -I/Users/rob/Downloads/
sage-3.0.rc0/local/include -g -O2 -Wall -Wpointer-arith -MT cipher.lo -
MD -MP -MF .deps/cipher.Tpo -c cipher.c -o cipher.o >/dev/null 2>&1
mv -f .deps/cipher.Tpo .deps/cipher.Plo
/bin/sh ../libtool --tag=CC --mode=compile gcc -DHAVE_CONFIG_H -I. -
I.. -I../src -I../src -I/Users/rob/Downloads/sage-3.0.rc0/local/
include -I/Users/rob/Downloads/sage-3.0.rc0/local/include -g -O2 -Wall
-Wpointer-arith -MT pubkey.lo -MD -MP -MF .deps/pubkey.Tpo -c -o
pubkey.lo pubkey.c
gcc -DHAVE_CONFIG_H -I. -I.. -I../src -I../src -I/Users/rob/
Downloads/sage-3.0.rc0/local/include -I/Users/rob/Downloads/
sage-3.0.rc0/local/include -g -O2 -Wall -Wpointer-arith -MT pubkey.lo -
MD -MP -MF .deps/pubkey.Tpo -c pubkey.c -fno-common -DPIC -o .libs/
pubkey.o
...
gcc -DHAVE_CONFIG_H -I. -I.. -I../src -I../src -I/Users/rob/
Downloads/sage-3.0.rc0/local/include -I/Users/rob/Downloads/
sage-3.0.rc0/local/include -g -O2 -Wall -Wpointer-arith -MT random.lo -
MD -MP -MF .deps/random.Tpo -c random.c -fno-common -DPIC -o .libs/
random.o
gcc -DHAVE_CONFIG_H -I. -I.. -I../src -I../src -I/Users/rob/
Downloads/sage-3.0.rc0/local/include -I/Users/rob/Downloads/
sage-3.0.rc0/local/include -g -O2 -Wall -Wpointer-arith -MT random.lo -
MD -MP -MF .deps/random.Tpo -c random.c -o random.o >/dev/null 2>&1
mv -f .deps/random.Tpo .deps/random.Plo
/bin/sh ../libtool --tag=CC --mode=compile gcc -DHAVE_CONFIG_H -I. -
I.. -I../src -I../src -I/Users/rob/Downloads/sage-3.0.rc0/local/
include -I/Users/rob/Downloads/sage-3.0.rc0/local/include -g -O2 -Wall
-Wpointer-arith -MT rndhw.lo -MD -MP -MF .deps/rndhw.Tpo -c -o
rndhw.lo rndhw.c
gcc -DHAVE_CONFIG_H -I. -I.. -I../src -I../src -I/Users/rob/
Downloads/sage-3.0.rc0/local/include -I/Users/rob/Downloads/
sage-3.0.rc0/local/include -g -O2 -Wall -Wpointer-arith -MT rndhw.lo -
MD -MP -MF .deps/rndhw.Tpo -c rndhw.c -fno-common -DPIC -o .libs/
rndhw.o
rndhw.c: In function ‘poll_padlock’:
rndhw.c:60: error: asm blocks not enabled, use `-fasm-blocks'
rndhw.c:60: error: expected identifier before ‘(’ token
rndhw.c:60: confused by earlier errors, bailing out
make[4]: *** [rndhw.lo] Error 1
make[3]: *** [all-recursive] Error 1
make[2]: *** [all] Error 2
failed to build libgcrypt
real 0m34.193s
user 0m17.394s
sys 0m14.601s
sage: An error occurred while installing libgcrypt-1.4.0.p1
Please email sage-devel http://groups.google.com/group/sage-devel
explaining the problem and send the relevant part of
of /Users/rob/Downloads/sage-3.0.rc0/install.log. Describe your
computer, operating system, etc.
If you want to try to fix the problem, yourself *don't* just cd to
/Users/rob/Downloads/sage-3.0.rc0/spkg/build/libgcrypt-1.4.0.p1 and
type 'make'.
Instead type "/Users/rob/Downloads/sage-3.0.rc0/sage -sh"
in order to set all environment variables correctly, then cd to
/Users/rob/Downloads/sage-3.0.rc0/spkg/build/libgcrypt-1.4.0.p1
(When you are done debugging, you can type "exit" to leave the
subshell.)
make[1]: *** [installed/libgcrypt-1.4.0.p1] Error 1
real 15m9.941s
user 11m47.843s
sys 2m35.299s
---------------------------------------------------------------
William Stein
> Hi,
>
> Brand new to Sage builds!
>
> On Mac OS 10.5.2, with:
>
> Robs-Intel:sage-3.0.rc0 rob$ gcc -v
> Using built-in specs.
> Target: i686-apple-darwin9
> Configured with: /var/tmp/gcc_42/gcc_42-5531~1/src/configure --disable-
> checking -enable-werror --prefix=/usr --mandir=/usr/share/man --enable-
> languages=c,objc,c++,obj-c++ --program-transform-name=/^[cg][^.-]*$/s/
> $/-4.2/ --with-gxx-include-dir=/usr/include/c++/4.0.0 --with-slibdir=/
> usr/lib --build=i686-apple-darwin9 --host=i686-apple-darwin9 --
> target=i686-apple-darwin9
> Thread model: posix
> gcc version 4.2.1 (Apple Inc. build 5531)
Which version of XCode is this? When did you get it?
I'm sure your problem is that we haven't ported Sage to work
with Apple's heavily hacked GCC-4.2.1.
I've been building all my OS X code on all platforms using
Apples GCC-4.0.1. I wasn't aware that they released GCC-4.2.1.
My guess is that numerous things will have to be fixed to get
Sage to build against APPLE-GCC-4.2.1. Your best bet
is to install a binary or install an older XCode (with GCC-4.0.1)
or wait until Sage supports GCC-4.2.1, which means 1-2 weeks
(since obviously I'll upgrade and have to support the newer compiler).
The compiler I use:
bsd:~ was$ gcc -v
Using built-in specs.
Target: i686-apple-darwin9
Configured with: /var/tmp/gcc/gcc-5465~16/src/configure
--disable-checking -enable-werror --prefix=/usr --mandir=/share/man
--enable-languages=c,objc,c++,obj-c++
--program-transform-name=/^[cg][^.-]*$/s/$/-4.0/
--with-gxx-include-dir=/include/c++/4.0.0 --with-slibdir=/usr/lib
--build=i686-apple-darwin9 --with-arch=apple --with-tune=generic
--host=i686-apple-darwin9 --target=i686-apple-darwin9
Thread model: posix
gcc version 4.0.1 (Apple Inc. build 5465)
bsd:~ was$
-- William
--
William Stein
Associate Professor of Mathematics
University of Washington
http://wstein.org
mabshoff
On Apr 21, 6:54 pm, "William Stein" <wst...@gmail.com> wrote:
> > Hi,
>
> > Brand new to Sage builds!
>
> > On Mac OS 10.5.2, with:
>
> > Robs-Intel:sage-3.0.rc0 rob$ gcc -v
> > Using built-in specs.
> > Target: i686-apple-darwin9
> > Configured with: /var/tmp/gcc_42/gcc_42-5531~1/src/configure --disable-
> > checking -enable-werror --prefix=/usr --mandir=/usr/share/man --enable-
> > languages=c,objc,c++,obj-c++ --program-transform-name=/^[cg][^.-]*$/s/
> > $/-4.2/ --with-gxx-include-dir=/usr/include/c++/4.0.0 --with-slibdir=/
> > usr/lib --build=i686-apple-darwin9 --host=i686-apple-darwin9 --
> > target=i686-apple-darwin9
> > Thread model: posix
> > gcc version 4.2.1 (Apple Inc. build 5531)
>
> ^^^^^^^^^^^^^^^^^^^
>
> Which version of XCode is this? When did you get it?
> I'm sure your problem is that we haven't ported Sage to work
> with Apple's heavily hacked GCC-4.2.1.
on VIA CPUs IIRC. So unless somebody did something they are no
supposed to it is something we can configure around.
> I've been building all my OS X code on all platforms using
> Apples GCC-4.0.1. I wasn't aware that they released GCC-4.2.1.
> My guess is that numerous things will have to be fixed to get
> Sage to build against APPLE-GCC-4.2.1. Your best bet
> is to install a binary or install an older XCode (with GCC-4.0.1)
> or wait until Sage supports GCC-4.2.1, which means 1-2 weeks
> (since obviously I'll upgrade and have to support the newer compiler).
in Sage 3.0.1.
Cheers,
Michael
Rob Goedman
site. It sure seems to compile an awful lot before hitting this.
I wonder how to make this work with R. R has these 100s of packages,
many of which need to be compiled when installed.
Right now, at least on the Mac, R-2-7 itself supports/contains 4
architectures ( ppc, x86 each 32 and 64 bits ) and the compilers for
c, c++ and fortran need to match for above package installations
(those that need to be compiled). Until recently there were disastrous
bugs in the compilers for 64 bits (and other issues with gcc4.0, e.g.
when fortran was involved).
Sage rc1 seems to include R-2.6.1, this week's R-2.7 release will
replace R-2.6.2.
I have tried gcc-4.0.1 but on my system it will hang forever (might be
related to the broken X11 that ships with Leopard):
gcc -bundle -undefined dynamic_lookup build/temp.macosx-10.3-i386-2.5/
Users/rob/Downloads/sage-3.0.rc1/spkg/build/python-2.5.2/src/Mac/
Modules/launch/_Launchmodule.o -L/Users/rob/Downloads/sage-3.0.rc1/
local/lib -L/usr/local/lib -o build/lib.macosx-10.3-i386-2.5/
_Launch.so -framework ApplicationServices
*** WARNING: renaming "_Launch" since importing it failed:
dlopen(build/lib.macosx-10.3-i386-2.5/_Launch.so, 2): Symbol not
found: __cg_png_create_info_struct
Referenced from: /System/Library/Frameworks/
ApplicationServices.framework/Versions/A/Frameworks/ImageIO.framework/
Versions/A/ImageIO
Expected in: /usr/X11R6/lib/libPng.dylib
building '_CG' extension
creating build/temp.macosx-10.3-i386-2.5/Users/rob/Downloads/
sage-3.0.rc1/spkg/build/python-2.5.2/src/Mac/Modules/cg
gcc -fno-strict-aliasing -Wno-long-double -no-cpp-precomp -mno-fused-
madd -DNDEBUG -g -fwrapv -O3 -Wall -Wstrict-prototypes -I. -I/Users/
rob/Downloads/sage-3.0.rc1/spkg/build/python-2.5.2/src/./Include -I/
Users/rob/Downloads/sage-3.0.rc1/spkg/build/python-2.5.2/src/./Mac/
Include -I/Users/rob/Downloads/sage-3.0.rc1/local/include -I. -
IInclude -I./Include -I/usr/local/include -I/Users/rob/Downloads/
sage-3.0.rc1/spkg/build/python-2.5.2/src/Include -I/Users/rob/
Downloads/sage-3.0.rc1/spkg/build/python-2.5.2/src -c /Users/rob/
Downloads/sage-3.0.rc1/spkg/build/python-2.5.2/src/Mac/Modules/cg/
_CGmodule.c -o build/temp.macosx-10.3-i386-2.5/Users/rob/Downloads/
sage-3.0.rc1/spkg/build/python-2.5.2/src/Mac/Modules/cg/_CGmodule.o -
Wno-deprecated-declarations
gcc -bundle -undefined dynamic_lookup build/temp.macosx-10.3-i386-2.5/
Users/rob/Downloads/sage-3.0.rc1/spkg/build/python-2.5.2/src/Mac/
Modules/cg/_CGmodule.o -L/Users/rob/Downloads/sage-3.0.rc1/local/lib -
L/usr/local/lib -o build/lib.macosx-10.3-i386-2.5/_CG.so -framework
ApplicationServices
and it will never finish this step.
My hope is that I can literally use either R.app to work with R and
its libraries and Sage notebooks using the same version of R and
libraries. Maybe the way Sage works with Mathematica?
As MIchael suggested, 3.0.1 stuff!
Rob
>> -Wpointer-arith -MT pubkey.lo -MD -MP -MF .deps/pubkey.Tpo -c -o
>> pubkey.lo pubkey.c
>> gcc -DHAVE_CONFIG_H -I. -I.. -I../src -I../src -I/Users/rob/
>> Downloads/sage-3.0.rc0/local/include -I/Users/rob/Downloads/
>> sage-3.0.rc0/local/include -g -O2 -Wall -Wpointer-arith -MT
>> pubkey.lo -
>> MD -MP -MF .deps/pubkey.Tpo -c pubkey.c -fno-common -DPIC -o .libs/
>> pubkey.o
>>
>> ...
>>
>> gcc -DHAVE_CONFIG_H -I. -I.. -I../src -I../src -I/Users/rob/
>> Downloads/sage-3.0.rc0/local/include -I/Users/rob/Downloads/
>> sage-3.0.rc0/local/include -g -O2 -Wall -Wpointer-arith -MT
>> random.lo -
>> MD -MP -MF .deps/random.Tpo -c random.c -fno-common -DPIC -o .libs/
>> random.o
>> gcc -DHAVE_CONFIG_H -I. -I.. -I../src -I../src -I/Users/rob/
>> Downloads/sage-3.0.rc0/local/include -I/Users/rob/Downloads/
>> sage-3.0.rc0/local/include -g -O2 -Wall -Wpointer-arith -MT
>> random.lo -
>> MD -MP -MF .deps/random.Tpo -c random.c -o random.o >/dev/null 2>&1
>> mv -f .deps/random.Tpo .deps/random.Plo
>> /bin/sh ../libtool --tag=CC --mode=compile gcc -DHAVE_CONFIG_H -
>> I. -
>> I.. -I../src -I../src -I/Users/rob/Downloads/sage-3.0.rc0/local/
>> include -I/Users/rob/Downloads/sage-3.0.rc0/local/include -g -O2 -
>> Wall
mabshoff
Hi Rob,
> This is XCode 3.0 with the updated gcc suite from the Apple developer
> site. It sure seems to compile an awful lot before hitting this.
>
> I wonder how to make this work with R. R has these 100s of packages,
> many of which need to be compiled when installed.
>
> Right now, at least on the Mac, R-2-7 itself supports/contains 4
> architectures ( ppc, x86 each 32 and 64 bits ) and the compilers for
> c, c++ and fortran need to match for above package installations
> (those that need to be compiled). Until recently there were disastrous
> bugs in the compilers for 64 bits (and other issues with gcc4.0, e.g.
> when fortran was involved).
>
> Sage rc1 seems to include R-2.6.1, this week's R-2.7 release will
> replace R-2.6.2.
version we are compiling ourselves in Sage since Apple's libPng leaves
much to be desired, i.e. a header file for example ;)
> and it will never finish this step.
>
> My hope is that I can literally use either R.app to work with R and
> its libraries and Sage notebooks using the same version of R and
> libraries. Maybe the way Sage works with Mathematica?
>
> As MIchael suggested, 3.0.1 stuff!
soon. The pad lock issue OSX should be fixed in the spkg at
http://sage.math.washington.edu/home/mabshoff/SPKG/libgcrypt-1.4.0.p2.spkg
This is now tracked at http://trac.sagemath.org/sage_trac/ticket/2993
> Rob
Rob: if you could verify that it fixes the issue it will be in 3.0.1.
We disabled padlock support on non-OSX. It seems that some version of
gcc 4.2 are slightly stupid and make configure believe that padlock
support is available on non-Via CPUs. Go figure.
Cheers,
Michael
William Stein
>
> This is XCode 3.0 with the updated gcc suite from the Apple developer
> site. It sure seems to compile an awful lot before hitting this.
How does one get the "updated gcc suite from the Apple Developer site"?
I just found http://developer.apple.com/tools/download/
I absolutely can't find gcc-4.2.1 anywhere on the apple website,
and it's definitely not included in XCode 3.0 as far as I can tell.
Thanks.
>
> I wonder how to make this work with R. R has these 100s of packages,
> many of which need to be compiled when installed.
>
> Right now, at least on the Mac, R-2-7 itself supports/contains 4
> architectures ( ppc, x86 each 32 and 64 bits ) and the compilers for
> c, c++ and fortran need to match for above package installations
> (those that need to be compiled). Until recently there were disastrous
> bugs in the compilers for 64 bits (and other issues with gcc4.0, e.g.
> when fortran was involved).
>
> Sage rc1 seems to include R-2.6.1, this week's R-2.7 release will
> replace R-2.6.2.
>
> I have tried gcc-4.0.1 but on my system it will hang forever (might be
> related to the broken X11 that ships with Leopard):
I've built Sage with gcc-4.0.1 many times on many flavors of OS X...
Yes, that is certainly going to be possible with Sage-3.0. It is one
of the main new features.