METATICKET Issues preventing 64-bit builds on various flavors of Solaris.

[sagetrac-drkirkby]

The following is list of the issues that are currently preventing Sage build 64-bit on either

• OpenSolaris (aka Solaris 11) on Intel/AMD x64. (e.g. 'disk')
• Solaris 10 on Intel/AMD (e.g. 'fulva')
• Solaris 10 on SPARC (e.g. 't2')

The list can be added as new problems are found, making what William calls a 'metaticket'

Please put

• "bug" if one knows, or might reasonable expect, it to cause a problem on a particular platform.
• "OK" if one knows the problem does not exist on that platform, or one reasonably expect it will not be an issue.
• "unknown" if one can not determine if it will be an issue.

Although OpenSolaris can run on the SPARC platform, very few people run it, so there is little to be gained by tracking issues on that platform. (David Kirkby does not have OpenSolaris installed on any of his SPARC systems)

Trac	Title	S10 SPARC	S10 x64	OpenSolaris x64	Notes
#7861	pynac not building on Open Solaris x64 (32-bit/64-bit mixup)	unknown	unknown	bug
#7864	libfplll tries to link 64-bit objects to 32-bit libstdc++.so	unknown	unknown	bug
#7982	sage_fortran is not working properly on Open Solaris x64 in 64 bit mode.	bug	bug	bug
#8086	numpy fails to build on Open Solaris x64 - 32 / 64-bit mixup	unknown	unknown	bug
#8089	ecl 9.10.2-20091105cvs.p1 fails to build on Open Solaris x64	unknown	unknown	bug
#9008	Update zlib to latest upstream, and clean up spkg-install	bug	unknown	bug
#9009	Mercurial is not building 64-bit with OpenSolaris with SAGE64=yes.	bug	bug	bug
#9022	python fails to build _socket on OpenSolaris x64, so ipython fails to build.	unknown	unknown	bug
#9023	ghmm needs compiler flag -m64 when SAGE64 is set to "yes"	bug	bug	bug
#9024	tachyon is buiding 32-bit on OpenSolaris x64 even when SAGE64 is set to "yes"	bug	bug	bug
#9025	PALP is building 32-bit on OpenSolaris - probably other platforms too.	bug	bug	bug
#9029	sympow is buiding 32-bit on OpenSolaris x64 even when SAGE64 is set to "yes"	bug	bug	bug
#9030	rubiks is building part 32-bit and part 64-bit on OpenSolaris x64.	bug	bug	bug
#9034	flintqs builds as 32-bit despite SAGE64=yes on OpenSolaris x64	bug	bug	bug
#9037	pynac fails to build on 64-bit OpenSolaris x64.	bug	bug	bug	Pynac 0.2 now in Sage. That must be fixed - see #8903 too
#9040	fatal relocation error installing R on OpenSolaris	unknown	unknown	bug	We many need to build R with SunStudio on OpenSolaris or Solaris 10 on x64
#9041	python fails to build _socket on OpenSolaris x64, so pygments fails to build.	unknown	unknown	bug
#9042	Cython fails to build on OpenSolaris x64	unknown	unknown	bug
#9043	lcalc failing to build on OpenSolaris x64.	bug	bug	bug
#9097	c_lib in Sage library fails to build on OpenSolaris x64	unknown	unknown	bug
#9098	gap buillds 32-bit on OpenSolaris when SAGE64=yes	bug	bug	bug
#9099	Maxima fails to build on OpenSolaris x64, though ECL does.	OK	bug	bug
#9100	scipy is probably building part 32-bit on OpenSolaris x64 when SAGE64=yes	unknown	unknown	bug	This is odd - see trac ticket
#9101	linbox reports "ERROR: BLAS not found!" on Solaris 10 SPARC (64-bit build).	bug	unknown	OK	It should use ATLAS, which mis-compiled as 32-bit
#9358	zn_poly passes all tests on on Solaris 10 64-bit SPARC, but fails to install	bug	bug	bug	positive review
#9397	Resolve corrupted patches to permit Singular to build on Solaris x86/x64	bug	unknown	OK	Should be fixed by update of Singular at #8059
#9399	Remove Sun-specific junk in rings/finite_rings/stdint.h	OK	bug	bug
#9508	Fix all ATLAS build problems on Solaris/OpenSolaris	bug	bug	bug
#9643	Force ECL to disable assembly code on Solaris 10 x86 as it does on OpenSolaris	OK	bug	OK	Positive review (OpenSolaris issue was resolved in #9474)

CC: @dimpase @jaapspies @sagetrac-mvngu @jhpalmieri @robertwb @williamstein

Component: porting: Solaris

Reviewer: Dima Pasechnik

Issue created by migration from https://trac.sagemath.org/ticket/9026

[jhpalmieri]

comment:24

As part of this project, we should update the "supported platforms" information, wherever it appears. For example, in the installation guide, it says

Complete compilation of Sage is currently not supported on Solaris

According to the "search_doc" function, the main occurrences of the string "Solaris" are in the file "devel/sage/doc/en/installation/source.rst", so that's the only part of the docs that we need to fix. The file SAGE_ROOT/README.txt says that Solaris on sparc is officially supported, but not Solaris on x86_64. I'm guessing that this is still accurate. There are probably assorted wiki pages about supported platforms, too. In any case, it would be good to maintain a single official list of supported platforms, instead of having conflicting information in different locations.

[mkoeppe]

comment:46

Should be closed as outdated.

[dimpase]

Reviewer: Dima Pasechnik

[fchapoton]

comment:48

agreed