Enable cephes on FreeBSD and disable it on Cygwin #9543

peterjeremy · 2010-07-18T19:57:13Z

FreeBSD does not currently have a full C99 libm and therefore also needs cephes. The attached patch enables cephes on FreeBSD, enables error checking and renames the long double gamma() functions in line with C99. The latter two components are applicable to Cygwin as well as FreeBSD.

Spkg at http://boxen.math.washington.edu/home/jpflori/cephes-2.8.p1.spkg

Additional comments for perusal in the future:

A number of areas still need work - in particular some of the cephes self tests have been removed and some of the other self tests fail (at least some of the failures reflect bugs in cephes). Suitable self-tests and a spkg-check script need to be developed. It's possible mpfr could be used as a reference library.

CC: @mwhansen @jpflori

Component: porting: BSD

Keywords: cephes spkg cygwin freebsd

Author: Peter Jeremy, Jean-Pierre Flori

Reviewer: Stephen Montgomery-Smith, Karl-Dieter Crisman

Merged: sage-5.6.rc0

Issue created by migration from https://trac.sagemath.org/ticket/9543

mwhansen · 2010-10-12T01:03:08Z

comment:2

I made an spkg at http://sage.math.washington.edu/home/mhansen/cephes-2.8.p0.spkg , but I get the following build error on Cygwin:


 /usr/lib/gcc/i686-pc-cygwin/4.3.4/collect2.exe -tsaware --wrap _Znwj --wrap _Znaj --wrap _ZdlPv --wrap _ZdaPv --wrap _ZnwjRKSt9nothrow_t --wrap _ZnajRKSt9nothrow_t --wrap _ZdlPvRKSt9nothrow_t --wrap _ZdaPvRKSt9nothrow_t -Bdynamic --dll-search-prefix=cyg -u ___register_frame_info -u ___deregister_frame_info -o mtst.exe /usr/lib/gcc/i686-pc-cygwin/4.3.4/../../../crt0.o /usr/lib/gcc/i686-pc-cygwin/4.3.4/crtbegin.o -L/home/mhansen/sage-4.5.3.alpha2/local/lib -L/home/mhansen/sage-4.5.3.alpha2/local/lib -L/usr/lib/gcc/i686-pc-cygwin/4.3.4 -L/usr/lib/gcc/i686-pc-cygwin/4.3.4 -L/usr/lib/gcc/i686-pc-cygwin/4.3.4/../../.. mtst.o cmplx.o clog.o cgamma.o stubs.o -lm -lgcc_s -lgcc -lgcc_eh -lcygwin -luser32 -lkernel32 -ladvapi32 -lshell32 -lgcc_s -lgcc -lgcc_eh /usr/lib/gcc/i686-pc-cygwin/4.3.4/crtend.o
gcc  -I. -g -O2 -Wall   -c -o cvect.o cvect.c
cvect.c: In function ‘main’:
cvect.c:130: warning: implicit declaration of function ‘memcmp’
cvect.c:223: warning: implicit declaration of function ‘exit’
cvect.c:223: warning: incompatible implicit declaration of built-in function ‘exit’
gcc -o cvect cvect.o clog.o cmplx.o stubs.o -lm
gcc  -I. -g -O2 -Wall -c dccalc.c
gcc -o dccalc dccalc.o libmc.a stubs.o -lm
gcc  -I. -g -O2 -Wall   -c -o mtstf.o mtstf.c
gcc -o mtstf mtstf.o cmplxf.o clogf.o cgammaf.o stubs.o -lm
make: *** No rule to make target `whitebxf.c', needed by `whitebxf.o'.  Stop.

peterjeremy · 2010-10-13T21:32:04Z

comment:3

The patches I added do not affect the build on Cygwin and I get exactly the same failure when building cephes-2.8.spkg (without any of my patches) on Cygwin. As I noted in my initial description, some of the cephes self-tests have been removed - whitebxf.c being one such.

jdemeyer · 2012-01-24T14:24:31Z

comment:4

Note that patches should now be applied by using patch instead of cp.

Also, it would be really nice to make it more uniform: when possible, use the same patches both for Cygwin and for FreeBSD.

kcrisman · 2012-01-24T14:32:21Z

comment:5

Replying to @jdemeyer:

Note that patches should now be applied by using patch instead of cp.

Also, it would be really nice to make it more uniform: when possible, use the same patches both for Cygwin and for FreeBSD.

Of course, though with Cygwin development halted for a bit that may not be as crucial.

kcrisman · 2012-01-31T01:58:57Z

comment:6

With respect to this patch right now, according to Stephen Montgomery-Smith, "It caused build errors in other sub-packages." However, it does seem that cephes is needed for ccosh and who knows what else.

kcrisman · 2012-01-31T01:59:12Z

comment:7

See this sage-devel thread.

sagetrac-stephen · 2012-04-14T03:30:57Z

comment:8

I could not get pjeremy's patch to work for FreeBSD. After some searching, I found out that the problem is that his patch attempts to link the cephes functions with /lib/libm.so. This is something that is not meant to work for dynamic libraries. Once a dynamic library has been created, apparently there is no mechanism for merging it with other dynamic libraries.

My solution was to slightly modify pjeremy's patch so that it creates a library libm_complex.so. The following patch should be applied to pjeremy's patch:

--- cephes-2.8	2012-04-14 01:39:13.000000000 +0000
+++ cephes-2.8	2012-04-14 01:38:51.000000000 +0000
@@ -1273,11 +1273,11 @@
 +# Intermediate (ar) libraries
 +LIBS=c9x-complex/libmc.a double/libmd.a ldouble/libml.a single/libmf.a
 +
-+all: libm.so
++all: libm_complex.so
 +
-+install: libm.so complex.h math.h
++install: libm_complex.so complex.h math.h
 +	${INSTALL} -C -m 644 complex.h math.h "${SAGE_LOCAL}/include"
-+	${INSTALL} -C -m 755 libm.so "${SAGE_LOCAL}/lib"
++	${INSTALL} -C -m 755 libm_complex.so "${SAGE_LOCAL}/lib"
 +
 +check:
 +	cd c9x-complex && ${MAKE} "CC=${CC}" check
@@ -1291,7 +1291,7 @@
 +#	TBD
 +
 +clean:
-+	rm -f libm.so syms.c99 syms.libm syms.wanted
++	rm -f libm_complex.so syms.c99 syms.libm syms.wanted
 +	cd c9x-complex && ${MAKE} clean
 +	cd double && ${MAKE} clean
 +	cd ldouble && ${MAKE} clean
@@ -1300,8 +1300,8 @@
 +# FreeBSD includes some but not all of the C99 maths functions.  Build
 +# a "new" libm.so that uses cephes functions to replace the missing ones
 +# (listed in syms.wanted) and then fallback to the base libm.so
-+libm.so: ${LIBS} syms.wanted
-+	${LD} -shared -o $@ $$(sed 's/^/-u /' syms.wanted) -L/usr/lib -lc -lm \
++libm_complex.so: ${LIBS} syms.wanted
++	${LD} -shared -o $@ $$(sed 's/^/-u /' syms.wanted) -L/usr/lib \
 +	   ${LIBS} -lgcc
 +
 +# List of symbols defined in the FreeBSD base libc.so and libm.so

Then I put a script in $SAGE_ROOT/local/bin called "cc" which is a wrapper around the cc I really want to use:

#!/usr/local/bin/bash

# Intersperse a "-lm_complex" before "-lm".

n=0
for i in "$@"; do
  if [ "x$i" = "x-lm" ]; then
    arg[$n]="-lm_complex"
    n=$((n+1))
    arg[$n]="-lm"
  else
    arg[$n]="$i"
  fi
  n=$((n+1))
done

# Some configure scripts invoke the compiler with the argument "-v", and if
# LDFLAGS are added to the arguments, this results in an error which
# ultimately stops the relevant package being built.

# Otherwise LDFLAGS needs to be added so that the linker knows where to find
# the dynamic libraries.

if [ $n = 1 -a "x${arg[0]}" = "x-v" ]; then
  exec /usr/local/bin/gcc46 "${arg[@]}"
else
  exec /usr/local/bin/gcc46 -Wl,-rpath=$SAGE_ROOT/local/lib  -Wl,-rpath=/usr/local/lib/gcc46 "${arg[@]}"
fi

jdemeyer · 2012-04-14T09:03:33Z

comment:9

Replying to @sagetrac-stephen:

I could not get pjeremy's patch to work for FreeBSD. After some searching, I found out that the problem is that his patch attempts to link the cephes functions with /lib/libm.so. This is something that is not meant to work for dynamic libraries.

Are you sure about this? Couldn't you link Cephes's libm against the system -lm (maybe you would have to rename the latter).

I'm willing to experiment with this if somebody could give me access to a FreeBSD box.

sagetrac-stephen · 2012-04-14T12:34:54Z

comment:10

Replying to @jdemeyer:

Replying to @sagetrac-stephen:

I could not get pjeremy's patch to work for FreeBSD. After some searching, I found out that the problem is that his patch attempts to link the cephes functions with /lib/libm.so. This is something that is not meant to work for dynamic libraries.

Are you sure about this? Couldn't you link Cephes's libm against the system -lm (maybe you would have to rename the latter).

I'm willing to experiment with this if somebody could give me access to a FreeBSD box.

No I am not sure about this. I received a private email from pjeremy, and did some more investigating. I discovered his patch doesn't work when I use the gcc46 compiler, which comes with the FreeBSD ports system, and is automatically invoked when you use fortran.

Let me think some more about this.

sagetrac-stephen · 2012-04-14T17:12:46Z

comment:11

Replying to @sagetrac-stephen:

No I am not sure about this. I received a private email from pjeremy, and did some more investigating. I discovered his patch doesn't work when I use the gcc46 compiler, which comes with the FreeBSD ports system, and is automatically invoked when you use fortran.

Let me think some more about this.

It looks like I found the problem. I was building under a chroot'ed environment. I have discovered that after I perform the

chroot /usr/jail

I need to do

sh /etc/rc.d/ldconfig start

It looks like I only need to do this one time.

Thanks for sticking with me on this guys. It looks like pjeremy's patch works, as is, with no changes.

kcrisman · 2012-06-20T19:19:49Z

Reviewer: Stephen Montgomery-Smith

kcrisman · 2012-06-20T19:19:49Z

Author: Peter Jeremy

kcrisman · 2012-06-20T19:19:49Z

comment:12

Great, we just have to make a new spkg for this. Probably we should change to patches instead of entire files copied over, though these are all new files so it's not as big of an issue.

kcrisman · 2012-12-18T16:54:06Z

comment:13

On #13806 this message came up. Relevant? Sorry that we still haven't made an spkg for this...

jpflori · 2012-12-19T10:56:57Z

comment:15

Spkg cleanup and only targetting FreeBSD at
http://boxen.math.washington.edu/home/jpflori/cephes-2.8.p0.spkg

kcrisman · 2013-01-04T15:32:59Z

Changed author from Peter Jeremy to Peter Jeremy, Jean-Pierre Flori

kcrisman · 2013-01-04T15:32:59Z

comment:16

Trivial error

if [ "$UNAME" != "FreeBSD" ]; then 
    echo "We only install the cephes library on Cygwin and FreeBSD." 
    exit 0 
fi

Anyway, I'm trying this on Cygwin now.

sagetrac-stephen · 2013-01-04T16:52:35Z

comment:17

Regarding testing the cephes library - I don't think it is fair to
expect them to give super accurate answers. I believe that the linux
library functions will fail in the same way. Same with OpenBSD.

I have been working with the FreeBSD people to get the C99 math
functions into FreeBSD. They are remarkably fussy. I have been working
on the complex arc-trig functions casin(h), cacos(h), catan(h), and I
have written some extremely well tested and verified code which is
listed here: http://www.math.missouri.edu/~stephen/software/#catrig

Nevertheless they still won't commit my programs until I have made some
style corrections, and I don't have time right now to do this.

The only other package I know of that implements these functions well
are the boost libraries. And even those were buggy (my bug fixes were
recently accepted by them).

Another person is working on clog. The real part of clog(z) is
particularly hard to implement in the case that |z| is close to one.

The linux and OpenBSD libraries totally disregard the issues that give
rise to the huge errors.

And for casinh and cacosh, even the mpc libraries are badly written in
that their only saving grace is that they keep increasing the number of
internal digits until they are sure that their answer is correct. (If
you wanted something like 200 bits of accuracy for cacosh, some inputs
can take minutes to compute.)

Anyway, all this is to say that you shouldn't hold the cephes libraries
up to any kind of high standard. Treat them as work arounds so that
ccosh and such like will actually be compiled into sage on FreeBSD,
rather that the answers should be particularly accurate.

kcrisman · 2013-01-04T17:09:59Z

comment:18

Replying to @sagetrac-stephen:

Regarding testing the cephes library - I don't think it is fair to
expect them to give super accurate answers. I believe that the linux
library functions will fail in the same way. Same with OpenBSD.

I'm not sure anyone was suggesting doing this, just to at least provide such functions on Cygwin.

Anyway, all this is to say that you shouldn't hold the cephes libraries
up to any kind of high standard. Treat them as work arounds so that
ccosh and such like will actually be compiled into sage on FreeBSD,
rather that the answers should be particularly accurate.

Hopefully that is all we are asking for! I'm going to change the description to point out that we really just want this.

So... would you mind checking (at your convenience) whether this spkg indeed does the same thing as the patches from before and at your port? Then if it works on Cygwin (i.e., if Cygwin doesn't need it) we can merge this and simplify things slightly.

sagetrac-stephen · 2013-01-04T20:43:11Z

comment:19

This new spkg failed to build rather early in the process:

Found package cephes-2.8.p0 in spkg/standard/cephes-2.8.p0.spkg
cephes-2.8.p0
====================================================
Extracting package /usr/home/stephen/sage-devel/work/sage-5.6.beta2/spkg/standard/cephes-2.8.p0.spkg
-rw-r--r--  1 stephen  staff  2514467 Jan  4 20:36 /usr/home/stephen/sage-devel/work/sage-5.6.beta2/spkg/standard/cephes-2.8.p0.spkg
Finished extraction
****************************************************
Host system:
FreeBSD wilberforce 8.3-STABLE FreeBSD 8.3-STABLE #0: Wed Jan  2 15:53:46 CST 2013     root@wilberforce:/usr/obj/usr/src/sys/GENERIC  amd64
****************************************************
C compiler: gcc
C compiler version:
Using built-in specs.
COLLECT_GCC=/usr/local/bin/gcc46
COLLECT_LTO_WRAPPER=/usr/local/libexec/gcc46/gcc/x86_64-portbld-freebsd8.3/4.6.3/lto-wrapper
Target: x86_64-portbld-freebsd8.3
Configured with: ./../gcc-4.6.3/configure --disable-bootstrap --disable-nls --libdir=/usr/local/lib/gcc46 --libexecdir=/usr/local/libexec/gcc46 --program-suffix=46 --with-as=/usr/local/bin/as --with-gmp=/usr/local --with-gxx-include-dir=/usr/local/lib/gcc46/include/c++/ --with-ld=/usr/local/bin/ld --with-libiconv-prefix=/usr/local --with-pkgversion='FreeBSD Ports Collection' --with-system-zlib --enable-languages=c,c++,objc,fortran,java --prefix=/usr/local --mandir=/usr/local/man --infodir=/usr/local/info/gcc46 --build=x86_64-portbld-freebsd8.3
Thread model: posix
gcc version 4.6.3 (FreeBSD Ports Collection)
****************************************************
Applying patches (if any)...
patching file c9x-complex/cgamma.c
patching file c9x-complex/cgammaf.c
patching file c9x-complex/cgammal.c
patching file c9x-complex/makefile
patching file double/makefile
patching file ldouble/gammal.c
patching file ldouble/makefile
patching file Makefile
./spkg-install: line 21: check_error: command not found
Building Cephes...
make[3]: Entering directory `/usr/home/stephen/sage-devel/work/sage-5.6.beta2/spkg/build/cephes-2.8.p0/src'
make[3]: *** No rule to make target `../patches/complex_bsd.h', needed by `c9x-complex/complex.h'.  Stop.
make[3]: Leaving directory `/usr/home/stephen/sage-devel/work/sage-5.6.beta2/spkg/build/cephes-2.8.p0/src'
Error - Failed to build Cephes ... exiting

real    0m0.025s
user    0m0.002s
sys     0m0.013s
************************************************************************
Error installing package cephes-2.8.p0
************************************************************************

kcrisman · 2013-01-07T14:46:25Z

Work Issues: typo in spkg-install, check_error not defined, complex_bsd.h

jpflori · 2013-01-07T15:12:31Z

comment:22

I've upped a new spkg which:

fixes the exit message,
fixes the use of check_error,
does not add math_bsd or complex_bsd as pjeremy's patch did not lead to their creation or installation, the bsd.patch and spkg-install I've produced should replicate the behavior of what you get with pjeremy's patch.

jpflori · 2013-01-07T15:12:31Z

Changed keywords from none to cephes spkg cygwin freebsd

jpflori · 2013-01-07T15:12:31Z

Changed work issues from typo in spkg-install, check_error not defined, complex_bsd.h to none

kcrisman · 2013-01-07T15:19:58Z

comment:23

This is fine from my point of view. If Stephen can check whether it performs as promised then we'll be set.

jpflori · 2013-01-07T15:23:34Z

comment:24

Replying to @kcrisman:

This is fine from my point of view. If Stephen can check whether it performs as promised then we'll be set.

Agreed!

Quite strangely, I don't have access to a FreeBSD box :) (although I vaguely remember trying to setup one once)

sagetrac-stephen · 2013-01-07T20:28:13Z

comment:25

It still fails with the "complex_bsd.h" no rule message.

I looked at Jeremy's patch, and it seemed to me that it did create math_bsd.h and complex_bsd.h. For example, it has lines like this:

--- cephes-2.8/patches/complex_bsd.h.orig       2010-07-26 08:25:54.654310990 +1000
+++ cephes-2.8/patches/complex_bsd.h    2010-07-26 08:25:54.658310309 +1000

This is a diff against a non-existent file, using the "-N" option to diff. The patch program creates a new file.

Makefile then copies this from cephes-2.8/patches to where-ever it is needed. I would probably change the patch so that it is created directly in the directory where it is needed.

sagetrac-stephen · 2013-01-07T20:29:56Z

comment:26

Oh, and the reason it worked so well with cygwin is because it didn't do anything.

kcrisman · 2013-01-07T21:12:11Z

comment:27

Oh, and the reason it worked so well with cygwin is because it didn't do anything.

Yes, understood.

jpflori · 2013-01-07T22:13:21Z

comment:28

Ok, I had a closer look at jeremy's patch and indeed the *_bsd.h file are needed and copied by the new makefile.

I did not create them at first because the bunch of bash functions called on cygwin explicitely copied them whereas on bsd we just have make / make install calls which got me confused.

I'll fix that tomorrow.

jpflori · 2013-01-08T14:17:28Z

Attachment: cephes-2.8.p0.diff.gz

Spkg diff, for review only.

jpflori · 2013-01-08T14:19:05Z

comment:29

Everything is hopefully fine now.

I've slightly modified Jeremy's approach to create (or rather modify as far as complex.h is concerned) the various headers while patching the src dir, rather than overwriting them with Makefile rules.

sagetrac-stephen · 2013-01-08T23:13:47Z

comment:30

The latest version built and worked just fine.

kcrisman · 2013-01-09T06:22:59Z

comment:31

And the spkg-install looks good to me. Assuming that you formed it correctly, this should be good to go in!

jdemeyer · 2013-01-09T09:04:47Z

comment:32

Sorry to be annoying but could you please document why we need cephes on FreeBSD? Preferably something should be mentioned in SPKG.txt. In the ticket description I only see

FreeBSD does not currently have a full C99 libm and therefore also needs cephes

which is very vague.

sagetrac-stephen · 2013-01-09T13:14:26Z

comment:33

FreeBSD's math library does not have all the mathematics functions described in Sections 7.3 and 7.12 of the C99 standard:

http://www.open-std.org/jtc1/sc22/wg14/www/docs/n1124.pdf

Specifically many of the complex and long double functions are missing in FreeBSD.

Cephes provides those functions that are missing from FreeBSD. The Makefiles have been modified so that it only adds those functions missing from the version of FreeBSD in which compilation is taking place.

jpflori · 2013-01-09T15:26:33Z

comment:34

New spkg with SPKG.txt including a note with what stephen posted.
As the change is minimal I'm leaving this as positive review.

jpflori · 2013-01-09T15:26:55Z

Spkg diff, for review only.

jdemeyer · 2013-01-12T08:52:00Z

comment:35

Attachment: cephes-2.8.p1.diff.gz

jdemeyer · 2013-01-12T08:52:00Z

Merged: sage-5.6.rc0

peterjeremy added this to the sage-5.6 milestone Jul 18, 2010

peterjeremy added c: porting: bsd labels Jul 18, 2010

peterjeremy self-assigned this Jul 18, 2010

This comment has been minimized.

Sign in to view

kcrisman changed the title ~~Enable cephes on FreeBSD~~ Enable cephes on FreeBSD and disable it on Cygwin Jan 4, 2013

kcrisman added the s: needs review label Jan 4, 2013

jpflori added s: needs review and removed s: needs work labels Jan 7, 2013

kcrisman added s: positive review and removed s: needs review labels Jan 9, 2013

This comment has been minimized.

Sign in to view

jdemeyer removed the s: positive review label Jan 12, 2013

jdemeyer closed this as completed Jan 12, 2013

peterjeremy mentioned this issue Jun 2, 2012

Fix cvxopt on FreeBSD #9601

Closed

sagetrac-stephen mannequin mentioned this issue Apr 14, 2012

Get Pynac functions to work on FreeBSD #12401

Closed

Enable cephes on FreeBSD and disable it on Cygwin #9543

Enable cephes on FreeBSD and disable it on Cygwin #9543

Comments

peterjeremy commented Jul 18, 2010

This comment has been minimized.

mwhansen commented Oct 12, 2010

peterjeremy commented Oct 13, 2010

jdemeyer commented Jan 24, 2012

kcrisman commented Jan 24, 2012

kcrisman commented Jan 31, 2012

kcrisman commented Jan 31, 2012

sagetrac-stephen mannequin commented Apr 14, 2012

jdemeyer commented Apr 14, 2012

sagetrac-stephen mannequin commented Apr 14, 2012

sagetrac-stephen mannequin commented Apr 14, 2012

kcrisman commented Jun 20, 2012

kcrisman commented Jun 20, 2012

kcrisman commented Jun 20, 2012

kcrisman commented Dec 18, 2012

jpflori commented Dec 19, 2012

This comment has been minimized.

kcrisman commented Jan 4, 2013

kcrisman commented Jan 4, 2013

sagetrac-stephen mannequin commented Jan 4, 2013

This comment has been minimized.

kcrisman commented Jan 4, 2013

sagetrac-stephen mannequin commented Jan 4, 2013

kcrisman commented Jan 7, 2013

jpflori commented Jan 7, 2013

jpflori commented Jan 7, 2013

jpflori commented Jan 7, 2013

kcrisman commented Jan 7, 2013

jpflori commented Jan 7, 2013

sagetrac-stephen mannequin commented Jan 7, 2013

sagetrac-stephen mannequin commented Jan 7, 2013

kcrisman commented Jan 7, 2013

jpflori commented Jan 7, 2013

jpflori commented Jan 8, 2013

jpflori commented Jan 8, 2013

sagetrac-stephen mannequin commented Jan 8, 2013

kcrisman commented Jan 9, 2013

jdemeyer commented Jan 9, 2013

sagetrac-stephen mannequin commented Jan 9, 2013

jpflori commented Jan 9, 2013

This comment has been minimized.

jpflori commented Jan 9, 2013

jdemeyer commented Jan 12, 2013

jdemeyer commented Jan 12, 2013