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

sagemath / sage

Main repository of SageMath

https://www.sagemath.org

Other

1.33k stars 453 forks source link

METATICKET Issues preventing 64-bit builds on various flavors of Solaris. #9026

Closed bac7d3ea-3f1b-4826-8464-f0b53d5e12d2 closed 4 years ago

bac7d3ea-3f1b-4826-8464-f0b53d5e12d2 commented 14 years ago

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

bac7d3ea-3f1b-4826-8464-f0b53d5e12d2 commented 14 years ago

Description changed:

--- 
+++ 
@@ -22,5 +22,6 @@
 |#9023|ghmm is buiding 32-bit on OpenSolaris x64 even when SAGE64 is set to "yes"|yes|yes|yes||
 |#9024|tachyon is buiding 32-bit on OpenSolaris x64 even when SAGE64 is set to "yes"|yes|yes|yes||
 |#9025|PALP is building 32-bit on OpenSolaris - probably other platforms too.|yes|yes|yes|I suspect this building 64-bit on OS X too, as there is nothing in the Makefile to force the -m64 flag on any platform at all|
+|#9029|sympow is buiding 32-bit on OpenSolaris x64 even when SAGE64 is set to "yes"|yes|yes|yes|I doubt this would build 64-bit on any platform where 32-bit was the default|

bac7d3ea-3f1b-4826-8464-f0b53d5e12d2 commented 14 years ago

Description changed:

--- 
+++ 
@@ -23,5 +23,6 @@
 |#9024|tachyon is buiding 32-bit on OpenSolaris x64 even when SAGE64 is set to "yes"|yes|yes|yes||
 |#9025|PALP is building 32-bit on OpenSolaris - probably other platforms too.|yes|yes|yes|I suspect this building 64-bit on OS X too, as there is nothing in the Makefile to force the -m64 flag on any platform at all|
 |#9029|sympow is buiding 32-bit on OpenSolaris x64 even when SAGE64 is set to "yes"|yes|yes|yes|I doubt this would build 64-bit on any platform where 32-bit was the default|
+|#9030|rubiks is building part 32-bit and part 64-bit on OpenSolaris x64.||yes|yes|yes|I doubt this would build 64-bit on any platform where 32-bit was the default|

bac7d3ea-3f1b-4826-8464-f0b53d5e12d2 commented 14 years ago

Description changed:

--- 
+++ 
@@ -23,6 +23,7 @@
 |#9024|tachyon is buiding 32-bit on OpenSolaris x64 even when SAGE64 is set to "yes"|yes|yes|yes||
 |#9025|PALP is building 32-bit on OpenSolaris - probably other platforms too.|yes|yes|yes|I suspect this building 64-bit on OS X too, as there is nothing in the Makefile to force the -m64 flag on any platform at all|
 |#9029|sympow is buiding 32-bit on OpenSolaris x64 even when SAGE64 is set to "yes"|yes|yes|yes|I doubt this would build 64-bit on any platform where 32-bit was the default|
-|#9030|rubiks is building part 32-bit and part 64-bit on OpenSolaris x64.||yes|yes|yes|I doubt this would build 64-bit on any platform where 32-bit was the default|
+|#9030|rubiks is building part 32-bit and part 64-bit on OpenSolaris x64.|yes|yes|yes|I doubt this would build 64-bit on any platform where 32-bit was the default|
+|#9033|Singular does not try to build 64-bit on OpenSolaris x64|yes|yes|yes|Singular does not even build fully as 32-bit, though it installs a few 32-bit object files in $SAGE_LOCAL/bin|

bac7d3ea-3f1b-4826-8464-f0b53d5e12d2 commented 14 years ago

Description changed:

--- 
+++ 
@@ -16,14 +16,15 @@
 |        |         |             |           |                    |         |
 |--------|---------|-------------|-----------|--------------------|---------|
 |**Trac**|**Title**|**S10 SPARC**|**S10 x64**| **OpenSolaris x64**|**Notes**|
-|#9008|Update zlib to latest upstream, and clean up spkg-install|unknown|unknown|yes|Needs testing on several platforms.|
-|#9009|Mercurial is not building 64-bit with OpenSolaris with SAGE64=yes.|yes|yes|yes||
-|#9022|python fails to build _socket on OpenSolaris x64, so ipython fails to build.|yes|unknown|yes||
+|#9008|Update zlib to latest upstream, and clean up spkg-install|yes|unknown|yes|Needs testing on several platforms.|
+|#9009|Mercurial is not building 64-bit with OpenSolaris with SAGE64=yes.|yes|yes|yes|Now has positive review|
+|#9022|python fails to build _socket on OpenSolaris x64, so ipython fails to build.|unknown|unknown|yes||
 |#9023|ghmm is buiding 32-bit on OpenSolaris x64 even when SAGE64 is set to "yes"|yes|yes|yes||
 |#9024|tachyon is buiding 32-bit on OpenSolaris x64 even when SAGE64 is set to "yes"|yes|yes|yes||
 |#9025|PALP is building 32-bit on OpenSolaris - probably other platforms too.|yes|yes|yes|I suspect this building 64-bit on OS X too, as there is nothing in the Makefile to force the -m64 flag on any platform at all|
 |#9029|sympow is buiding 32-bit on OpenSolaris x64 even when SAGE64 is set to "yes"|yes|yes|yes|I doubt this would build 64-bit on any platform where 32-bit was the default|
 |#9030|rubiks is building part 32-bit and part 64-bit on OpenSolaris x64.|yes|yes|yes|I doubt this would build 64-bit on any platform where 32-bit was the default|
 |#9033|Singular does not try to build 64-bit on OpenSolaris x64|yes|yes|yes|Singular does not even build fully as 32-bit, though it installs a few 32-bit object files in $SAGE_LOCAL/bin|
+|#9034|flintqs builds as 32-bit despite SAGE64=yes on OpenSolaris x64|yes|yes|yes|A makefile included in flitqs for 64-bit OS X will probably work on OpenSolaris too. It needs checking.|

bac7d3ea-3f1b-4826-8464-f0b53d5e12d2 commented 14 years ago

Description changed:

--- 
+++ 
@@ -26,5 +26,6 @@
 |#9030|rubiks is building part 32-bit and part 64-bit on OpenSolaris x64.|yes|yes|yes|I doubt this would build 64-bit on any platform where 32-bit was the default|
 |#9033|Singular does not try to build 64-bit on OpenSolaris x64|yes|yes|yes|Singular does not even build fully as 32-bit, though it installs a few 32-bit object files in $SAGE_LOCAL/bin|
 |#9034|flintqs builds as 32-bit despite SAGE64=yes on OpenSolaris x64|yes|yes|yes|A makefile included in flitqs for 64-bit OS X will probably work on OpenSolaris too. It needs checking.|
+|#9037|pynac fails to build on 64-bit OpenSolaris x64.|yes|yes|yes|The usual problem of 64-bit build restricted to OS X|

bac7d3ea-3f1b-4826-8464-f0b53d5e12d2 commented 14 years ago

Description changed:

--- 
+++ 
@@ -9,7 +9,7 @@
 Please put 
 * "yes" if the trac ticket is known, or one might reasonably expect it to cause a problem on a particular platform.
 * "no" if one knows the problem does not exist on that platform, or one reasonably expect it will not be an issue. 
-* "unknow" if one can not determine if it will 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)

@@ -27,5 +27,6 @@
 |#9033|Singular does not try to build 64-bit on OpenSolaris x64|yes|yes|yes|Singular does not even build fully as 32-bit, though it installs a few 32-bit object files in $SAGE_LOCAL/bin|
 |#9034|flintqs builds as 32-bit despite SAGE64=yes on OpenSolaris x64|yes|yes|yes|A makefile included in flitqs for 64-bit OS X will probably work on OpenSolaris too. It needs checking.|
 |#9037|pynac fails to build on 64-bit OpenSolaris x64.|yes|yes|yes|The usual problem of 64-bit build restricted to OS X|
+|#9042|Cython fails to build on OpenSolaris x64|unknown|unknown|yes||

bac7d3ea-3f1b-4826-8464-f0b53d5e12d2 commented 14 years ago

Description changed:

--- 
+++ 
@@ -16,17 +16,18 @@
 |        |         |             |           |                    |         |
 |--------|---------|-------------|-----------|--------------------|---------|
 |**Trac**|**Title**|**S10 SPARC**|**S10 x64**| **OpenSolaris x64**|**Notes**|
+|#7982|sage_fortran is not working properly on Open Solaris x64 in 64 bit mode.|yes|yes|yes|**needs review**|
 |#9008|Update zlib to latest upstream, and clean up spkg-install|yes|unknown|yes|Needs testing on several platforms.|
 |#9009|Mercurial is not building 64-bit with OpenSolaris with SAGE64=yes.|yes|yes|yes|Now has positive review|
 |#9022|python fails to build _socket on OpenSolaris x64, so ipython fails to build.|unknown|unknown|yes||
-|#9023|ghmm is buiding 32-bit on OpenSolaris x64 even when SAGE64 is set to "yes"|yes|yes|yes||
-|#9024|tachyon is buiding 32-bit on OpenSolaris x64 even when SAGE64 is set to "yes"|yes|yes|yes||
-|#9025|PALP is building 32-bit on OpenSolaris - probably other platforms too.|yes|yes|yes|I suspect this building 64-bit on OS X too, as there is nothing in the Makefile to force the -m64 flag on any platform at all|
-|#9029|sympow is buiding 32-bit on OpenSolaris x64 even when SAGE64 is set to "yes"|yes|yes|yes|I doubt this would build 64-bit on any platform where 32-bit was the default|
+|#9023|ghmm is buiding 32-bit on OpenSolaris x64 even when SAGE64 is set to "yes"|yes|yes|yes|**needs review**|
+|#9024|tachyon is buiding 32-bit on OpenSolaris x64 even when SAGE64 is set to "yes"|yes|yes|yes|**needs review**|
+|#9025|PALP is building 32-bit on OpenSolaris - probably other platforms too.|yes|yes|yes|I suspect this building 64-bit on OS X too, as there is nothing in the Makefile to force the -m64 flag on any platform at all. **needs review too**|
+|#9029|sympow is buiding 32-bit on OpenSolaris x64 even when SAGE64 is set to "yes"|yes|yes|yes|**needs review**|
 |#9030|rubiks is building part 32-bit and part 64-bit on OpenSolaris x64.|yes|yes|yes|I doubt this would build 64-bit on any platform where 32-bit was the default|
 |#9033|Singular does not try to build 64-bit on OpenSolaris x64|yes|yes|yes|Singular does not even build fully as 32-bit, though it installs a few 32-bit object files in $SAGE_LOCAL/bin|
-|#9034|flintqs builds as 32-bit despite SAGE64=yes on OpenSolaris x64|yes|yes|yes|A makefile included in flitqs for 64-bit OS X will probably work on OpenSolaris too. It needs checking.|
+|#9034|flintqs builds as 32-bit despite SAGE64=yes on OpenSolaris x64|yes|yes|yes|**needs review**|
 |#9037|pynac fails to build on 64-bit OpenSolaris x64.|yes|yes|yes|The usual problem of 64-bit build restricted to OS X|
 |#9042|Cython fails to build on OpenSolaris x64|unknown|unknown|yes||
-
-
+|#9043|lcalc failing to build on OpenSolaris x64.|unknown|unknown|yes||
+|#9040|fatal relocation error installing R on OpenSolaris|unknown|unknown|yes|

bac7d3ea-3f1b-4826-8464-f0b53d5e12d2 commented 14 years ago

Description changed:

--- 
+++ 
@@ -17,7 +17,7 @@
 |--------|---------|-------------|-----------|--------------------|---------|
 |**Trac**|**Title**|**S10 SPARC**|**S10 x64**| **OpenSolaris x64**|**Notes**|
 |#7982|sage_fortran is not working properly on Open Solaris x64 in 64 bit mode.|yes|yes|yes|**needs review**|
-|#9008|Update zlib to latest upstream, and clean up spkg-install|yes|unknown|yes|Needs testing on several platforms.|
+|#9008|Update zlib to latest upstream, and clean up spkg-install|yes|unknown|yes|**needs review**|
 |#9009|Mercurial is not building 64-bit with OpenSolaris with SAGE64=yes.|yes|yes|yes|Now has positive review|
 |#9022|python fails to build _socket on OpenSolaris x64, so ipython fails to build.|unknown|unknown|yes||
 |#9023|ghmm is buiding 32-bit on OpenSolaris x64 even when SAGE64 is set to "yes"|yes|yes|yes|**needs review**|
@@ -29,5 +29,5 @@
 |#9034|flintqs builds as 32-bit despite SAGE64=yes on OpenSolaris x64|yes|yes|yes|**needs review**|
 |#9037|pynac fails to build on 64-bit OpenSolaris x64.|yes|yes|yes|The usual problem of 64-bit build restricted to OS X|
 |#9042|Cython fails to build on OpenSolaris x64|unknown|unknown|yes||
-|#9043|lcalc failing to build on OpenSolaris x64.|unknown|unknown|yes||
+|#9043|lcalc failing to build on OpenSolaris x64.|yes|yes|yes|**needs review**|
 |#9040|fatal relocation error installing R on OpenSolaris|unknown|unknown|yes|

bac7d3ea-3f1b-4826-8464-f0b53d5e12d2 commented 14 years ago

Description changed:

--- 
+++ 
@@ -16,18 +16,27 @@
 |        |         |             |           |                    |         |
 |--------|---------|-------------|-----------|--------------------|---------|
 |**Trac**|**Title**|**S10 SPARC**|**S10 x64**| **OpenSolaris x64**|**Notes**|
+|#7864|libfplll tries to link 64-bit objects to 32-bit libstdc++.so|unknown|unknown|yes||
 |#7982|sage_fortran is not working properly on Open Solaris x64 in 64 bit mode.|yes|yes|yes|**needs review**|
+|#8089|ecl 9.10.2-20091105cvs.p1 faiils to build on Open Solaris x64|unknown|unknown|yes|**needs review**|
 |#9008|Update zlib to latest upstream, and clean up spkg-install|yes|unknown|yes|**needs review**|
 |#9009|Mercurial is not building 64-bit with OpenSolaris with SAGE64=yes.|yes|yes|yes|Now has positive review|
-|#9022|python fails to build _socket on OpenSolaris x64, so ipython fails to build.|unknown|unknown|yes||
-|#9023|ghmm is buiding 32-bit on OpenSolaris x64 even when SAGE64 is set to "yes"|yes|yes|yes|**needs review**|
+|#9022|python fails to build _socket on OpenSolaris x64, so ipython fails to build.|unknown|unknown|yes|fixed by #9041 which **needs review**|
+|#9023|ghmm is buiding 32-bit on OpenSolaris x64 even when SAGE64 is set to "yes"|yes|yes|yes||
 |#9024|tachyon is buiding 32-bit on OpenSolaris x64 even when SAGE64 is set to "yes"|yes|yes|yes|**needs review**|
 |#9025|PALP is building 32-bit on OpenSolaris - probably other platforms too.|yes|yes|yes|I suspect this building 64-bit on OS X too, as there is nothing in the Makefile to force the -m64 flag on any platform at all. **needs review too**|
 |#9029|sympow is buiding 32-bit on OpenSolaris x64 even when SAGE64 is set to "yes"|yes|yes|yes|**needs review**|
-|#9030|rubiks is building part 32-bit and part 64-bit on OpenSolaris x64.|yes|yes|yes|I doubt this would build 64-bit on any platform where 32-bit was the default|
+|#9030|rubiks is building part 32-bit and part 64-bit on OpenSolaris x64.|yes|yes|yes|**needs review**|
 |#9033|Singular does not try to build 64-bit on OpenSolaris x64|yes|yes|yes|Singular does not even build fully as 32-bit, though it installs a few 32-bit object files in $SAGE_LOCAL/bin|
 |#9034|flintqs builds as 32-bit despite SAGE64=yes on OpenSolaris x64|yes|yes|yes|**needs review**|
-|#9037|pynac fails to build on 64-bit OpenSolaris x64.|yes|yes|yes|The usual problem of 64-bit build restricted to OS X|
-|#9042|Cython fails to build on OpenSolaris x64|unknown|unknown|yes||
+|#9037|pynac fails to build on 64-bit OpenSolaris x64.|yes|yes|yes|Pynac 0.2 now in Sage. That must be fixed - see #8903 too|
+|#9041|python fails to build _socket on OpenSolaris x64, so pygments fails to build.|unknown|unknown|yes|**needs review**|
+|#9042|Cython fails to build on OpenSolaris x64|unknown|unknown|yes|Positive review|
 |#9043|lcalc failing to build on OpenSolaris x64.|yes|yes|yes|**needs review**|
 |#9040|fatal relocation error installing R on OpenSolaris|unknown|unknown|yes|
+|#9097|c_lib in Sage library fails to build on OpenSolaris x64|unknown|unknown|yes|Could be tricky - SCons problem, and I dislike SCons|
+|#9098|gap buillds 32-bit on OpenSolaris when SAGE64=yes|yes|yes|yes|The -m64 option is not added on any platform|
+|#9099|Maxima fails to build on OpenSolaris x64, though ECL does.|unknown|unknown|yes||
+|#9100|scipy is probably building part 32-bit on OpenSolaris x64 when SAGE64=yes|unknown|unknown|yes|This is odd - see trac ticket|
+
+

bac7d3ea-3f1b-4826-8464-f0b53d5e12d2 commented 14 years ago

Description changed:

--- 
+++ 
@@ -18,7 +18,7 @@
 |**Trac**|**Title**|**S10 SPARC**|**S10 x64**| **OpenSolaris x64**|**Notes**|
 |#7864|libfplll tries to link 64-bit objects to 32-bit libstdc++.so|unknown|unknown|yes||
 |#7982|sage_fortran is not working properly on Open Solaris x64 in 64 bit mode.|yes|yes|yes|**needs review**|
-|#8089|ecl 9.10.2-20091105cvs.p1 faiils to build on Open Solaris x64|unknown|unknown|yes|**needs review**|
+|#8089|ecl 9.10.2-20091105cvs.p1 faiils to build on Open Solaris x64|unknown|unknown|yes|Solved by #8951, which is a later version of ECL|
 |#9008|Update zlib to latest upstream, and clean up spkg-install|yes|unknown|yes|**needs review**|
 |#9009|Mercurial is not building 64-bit with OpenSolaris with SAGE64=yes.|yes|yes|yes|Now has positive review|
 |#9022|python fails to build _socket on OpenSolaris x64, so ipython fails to build.|unknown|unknown|yes|fixed by #9041 which **needs review**|

bac7d3ea-3f1b-4826-8464-f0b53d5e12d2 commented 14 years ago

Description changed:

--- 
+++ 
@@ -38,5 +38,4 @@
 |#9098|gap buillds 32-bit on OpenSolaris when SAGE64=yes|yes|yes|yes|The -m64 option is not added on any platform|
 |#9099|Maxima fails to build on OpenSolaris x64, though ECL does.|unknown|unknown|yes||
 |#9100|scipy is probably building part 32-bit on OpenSolaris x64 when SAGE64=yes|unknown|unknown|yes|This is odd - see trac ticket|
-
-
+|#9101|linbox reports "ERROR: BLAS not found!" on Solaris 10 SPARC (64-bit build).|yes|unknown|no|I believe spkg/standard/deps is missing the fact linbox depends on BLAS|

bac7d3ea-3f1b-4826-8464-f0b53d5e12d2 commented 14 years ago

Description changed:

--- 
+++ 
@@ -16,6 +16,7 @@
 |        |         |             |           |                    |         |
 |--------|---------|-------------|-----------|--------------------|---------|
 |**Trac**|**Title**|**S10 SPARC**|**S10 x64**| **OpenSolaris x64**|**Notes**|
+|#7861|pynac not building on Open Solaris x64 (32-bit/64-bit mixup)|unknown|unknown|yes||
 |#7864|libfplll tries to link 64-bit objects to 32-bit libstdc++.so|unknown|unknown|yes||
 |#7982|sage_fortran is not working properly on Open Solaris x64 in 64 bit mode.|yes|yes|yes|**needs review**|
 |#8089|ecl 9.10.2-20091105cvs.p1 faiils to build on Open Solaris x64|unknown|unknown|yes|Solved by #8951, which is a later version of ECL|

bac7d3ea-3f1b-4826-8464-f0b53d5e12d2 commented 14 years ago

Description changed:

--- 
+++ 
@@ -21,21 +21,21 @@
 |#7982|sage_fortran is not working properly on Open Solaris x64 in 64 bit mode.|yes|yes|yes|**needs review**|
 |#8089|ecl 9.10.2-20091105cvs.p1 faiils to build on Open Solaris x64|unknown|unknown|yes|Solved by #8951, which is a later version of ECL|
 |#9008|Update zlib to latest upstream, and clean up spkg-install|yes|unknown|yes|**needs review**|
-|#9009|Mercurial is not building 64-bit with OpenSolaris with SAGE64=yes.|yes|yes|yes|Now has positive review|
+|#9009|Mercurial is not building 64-bit with OpenSolaris with SAGE64=yes.|yes|yes|yes|Now fixed|
 |#9022|python fails to build _socket on OpenSolaris x64, so ipython fails to build.|unknown|unknown|yes|fixed by #9041 which **needs review**|
 |#9023|ghmm is buiding 32-bit on OpenSolaris x64 even when SAGE64 is set to "yes"|yes|yes|yes||
-|#9024|tachyon is buiding 32-bit on OpenSolaris x64 even when SAGE64 is set to "yes"|yes|yes|yes|**needs review**|
-|#9025|PALP is building 32-bit on OpenSolaris - probably other platforms too.|yes|yes|yes|I suspect this building 64-bit on OS X too, as there is nothing in the Makefile to force the -m64 flag on any platform at all. **needs review too**|
+|#9024|tachyon is buiding 32-bit on OpenSolaris x64 even when SAGE64 is set to "yes"|yes|yes|yes|Now fixed|
+|#9025|PALP is building 32-bit on OpenSolaris - probably other platforms too.|yes|yes|yes|I suspect this building 64-bit on OS X too, as there is nothing in the Makefile to force the -m64 flag on any platform at all. **needs re-basing**|
 |#9029|sympow is buiding 32-bit on OpenSolaris x64 even when SAGE64 is set to "yes"|yes|yes|yes|**needs review**|
-|#9030|rubiks is building part 32-bit and part 64-bit on OpenSolaris x64.|yes|yes|yes|**needs review**|
+|#9030|rubiks is building part 32-bit and part 64-bit on OpenSolaris x64.|yes|yes|yes|positive review|
 |#9033|Singular does not try to build 64-bit on OpenSolaris x64|yes|yes|yes|Singular does not even build fully as 32-bit, though it installs a few 32-bit object files in $SAGE_LOCAL/bin|
 |#9034|flintqs builds as 32-bit despite SAGE64=yes on OpenSolaris x64|yes|yes|yes|**needs review**|
 |#9037|pynac fails to build on 64-bit OpenSolaris x64.|yes|yes|yes|Pynac 0.2 now in Sage. That must be fixed - see #8903 too|
-|#9041|python fails to build _socket on OpenSolaris x64, so pygments fails to build.|unknown|unknown|yes|**needs review**|
+|#9041|python fails to build _socket on OpenSolaris x64, so pygments fails to build.|unknown|unknown|yes|positive review|
 |#9042|Cython fails to build on OpenSolaris x64|unknown|unknown|yes|Positive review|
 |#9043|lcalc failing to build on OpenSolaris x64.|yes|yes|yes|**needs review**|
 |#9040|fatal relocation error installing R on OpenSolaris|unknown|unknown|yes|
-|#9097|c_lib in Sage library fails to build on OpenSolaris x64|unknown|unknown|yes|Could be tricky - SCons problem, and I dislike SCons|
+|#9097|c_lib in Sage library fails to build on OpenSolaris x64|unknown|unknown|yes|Looks easy - despite it's SCons!|
 |#9098|gap buillds 32-bit on OpenSolaris when SAGE64=yes|yes|yes|yes|The -m64 option is not added on any platform|
 |#9099|Maxima fails to build on OpenSolaris x64, though ECL does.|unknown|unknown|yes||
 |#9100|scipy is probably building part 32-bit on OpenSolaris x64 when SAGE64=yes|unknown|unknown|yes|This is odd - see trac ticket|

bac7d3ea-3f1b-4826-8464-f0b53d5e12d2 commented 14 years ago

Description changed:

--- 
+++ 
@@ -18,25 +18,25 @@
 |**Trac**|**Title**|**S10 SPARC**|**S10 x64**| **OpenSolaris x64**|**Notes**|
 |#7861|pynac not building on Open Solaris x64 (32-bit/64-bit mixup)|unknown|unknown|yes||
 |#7864|libfplll tries to link 64-bit objects to 32-bit libstdc++.so|unknown|unknown|yes||
-|#7982|sage_fortran is not working properly on Open Solaris x64 in 64 bit mode.|yes|yes|yes|**needs review**|
-|#8089|ecl 9.10.2-20091105cvs.p1 faiils to build on Open Solaris x64|unknown|unknown|yes|Solved by #8951, which is a later version of ECL|
-|#9008|Update zlib to latest upstream, and clean up spkg-install|yes|unknown|yes|**needs review**|
+|#7982|sage_fortran is not working properly on Open Solaris x64 in 64 bit mode.|yes|yes|yes|**positive review**|
+|#8089|ecl 9.10.2-20091105cvs.p1 faiils to build on Open Solaris x64|unknown|unknown|yes|Solved by #8951, which has positive review.|
+|#9008|Update zlib to latest upstream, and clean up spkg-install|yes|unknown|yes|**positive review**|
 |#9009|Mercurial is not building 64-bit with OpenSolaris with SAGE64=yes.|yes|yes|yes|Now fixed|
-|#9022|python fails to build _socket on OpenSolaris x64, so ipython fails to build.|unknown|unknown|yes|fixed by #9041 which **needs review**|
-|#9023|ghmm is buiding 32-bit on OpenSolaris x64 even when SAGE64 is set to "yes"|yes|yes|yes||
+|#9022|python fails to build _socket on OpenSolaris x64, so ipython fails to build.|unknown|unknown|yes|fixed by #9041 which has positive review|
+|#9023|ghmm is buiding 32-bit on OpenSolaris x64 even when SAGE64 is set to "yes"|yes|yes|yes|needs work|
 |#9024|tachyon is buiding 32-bit on OpenSolaris x64 even when SAGE64 is set to "yes"|yes|yes|yes|Now fixed|
 |#9025|PALP is building 32-bit on OpenSolaris - probably other platforms too.|yes|yes|yes|I suspect this building 64-bit on OS X too, as there is nothing in the Makefile to force the -m64 flag on any platform at all. **needs re-basing**|
-|#9029|sympow is buiding 32-bit on OpenSolaris x64 even when SAGE64 is set to "yes"|yes|yes|yes|**needs review**|
+|#9029|sympow is buiding 32-bit on OpenSolaris x64 even when SAGE64 is set to "yes"|yes|yes|yes|Fixed|
 |#9030|rubiks is building part 32-bit and part 64-bit on OpenSolaris x64.|yes|yes|yes|positive review|
 |#9033|Singular does not try to build 64-bit on OpenSolaris x64|yes|yes|yes|Singular does not even build fully as 32-bit, though it installs a few 32-bit object files in $SAGE_LOCAL/bin|
-|#9034|flintqs builds as 32-bit despite SAGE64=yes on OpenSolaris x64|yes|yes|yes|**needs review**|
+|#9034|flintqs builds as 32-bit despite SAGE64=yes on OpenSolaris x64|yes|yes|yes|Positive review|
 |#9037|pynac fails to build on 64-bit OpenSolaris x64.|yes|yes|yes|Pynac 0.2 now in Sage. That must be fixed - see #8903 too|
 |#9041|python fails to build _socket on OpenSolaris x64, so pygments fails to build.|unknown|unknown|yes|positive review|
 |#9042|Cython fails to build on OpenSolaris x64|unknown|unknown|yes|Positive review|
-|#9043|lcalc failing to build on OpenSolaris x64.|yes|yes|yes|**needs review**|
-|#9040|fatal relocation error installing R on OpenSolaris|unknown|unknown|yes|
+|#9043|lcalc failing to build on OpenSolaris x64.|yes|yes|yes|Positive review|
+|#9040|fatal relocation error installing R on OpenSolaris|unknown|unknown|yes|We many need to build R with SunStudio on OpenSolaris or Solaris 10 on x64|
 |#9097|c_lib in Sage library fails to build on OpenSolaris x64|unknown|unknown|yes|Looks easy - despite it's SCons!|
-|#9098|gap buillds 32-bit on OpenSolaris when SAGE64=yes|yes|yes|yes|The -m64 option is not added on any platform|
+|#9098|gap buillds 32-bit on OpenSolaris when SAGE64=yes|yes|yes|yes|The -m64 option is not added on any platform. Should be easy to fix.|
 |#9099|Maxima fails to build on OpenSolaris x64, though ECL does.|unknown|unknown|yes||
 |#9100|scipy is probably building part 32-bit on OpenSolaris x64 when SAGE64=yes|unknown|unknown|yes|This is odd - see trac ticket|
-|#9101|linbox reports "ERROR: BLAS not found!" on Solaris 10 SPARC (64-bit build).|yes|unknown|no|I believe spkg/standard/deps is missing the fact linbox depends on BLAS|
+|#9101|linbox reports "ERROR: BLAS not found!" on Solaris 10 SPARC (64-bit build).|yes|unknown|no|It should use ATLAS|

bac7d3ea-3f1b-4826-8464-f0b53d5e12d2 commented 14 years ago

Description changed:

--- 
+++ 
@@ -28,7 +28,7 @@
 |#9025|PALP is building 32-bit on OpenSolaris - probably other platforms too.|yes|yes|yes|I suspect this building 64-bit on OS X too, as there is nothing in the Makefile to force the -m64 flag on any platform at all. **needs re-basing**|
 |#9029|sympow is buiding 32-bit on OpenSolaris x64 even when SAGE64 is set to "yes"|yes|yes|yes|Fixed|
 |#9030|rubiks is building part 32-bit and part 64-bit on OpenSolaris x64.|yes|yes|yes|positive review|
-|#9033|Singular does not try to build 64-bit on OpenSolaris x64|yes|yes|yes|Singular does not even build fully as 32-bit, though it installs a few 32-bit object files in $SAGE_LOCAL/bin|
+|#9033|Singular does not try to build 64-bit on OpenSolaris x64|yes|yes|yes|#9160 does not solve it, but goes some way towards a solution|
 |#9034|flintqs builds as 32-bit despite SAGE64=yes on OpenSolaris x64|yes|yes|yes|Positive review|
 |#9037|pynac fails to build on 64-bit OpenSolaris x64.|yes|yes|yes|Pynac 0.2 now in Sage. That must be fixed - see #8903 too|
 |#9041|python fails to build _socket on OpenSolaris x64, so pygments fails to build.|unknown|unknown|yes|positive review|

bac7d3ea-3f1b-4826-8464-f0b53d5e12d2 commented 14 years ago

Description changed:

--- 
+++ 
@@ -25,7 +25,7 @@
 |#9022|python fails to build _socket on OpenSolaris x64, so ipython fails to build.|unknown|unknown|yes|fixed by #9041 which has positive review|
 |#9023|ghmm is buiding 32-bit on OpenSolaris x64 even when SAGE64 is set to "yes"|yes|yes|yes|needs work|
 |#9024|tachyon is buiding 32-bit on OpenSolaris x64 even when SAGE64 is set to "yes"|yes|yes|yes|Now fixed|
-|#9025|PALP is building 32-bit on OpenSolaris - probably other platforms too.|yes|yes|yes|I suspect this building 64-bit on OS X too, as there is nothing in the Makefile to force the -m64 flag on any platform at all. **needs re-basing**|
+|#9025|PALP is building 32-bit on OpenSolaris - probably other platforms too.|yes|yes|yes|Positive review|
 |#9029|sympow is buiding 32-bit on OpenSolaris x64 even when SAGE64 is set to "yes"|yes|yes|yes|Fixed|
 |#9030|rubiks is building part 32-bit and part 64-bit on OpenSolaris x64.|yes|yes|yes|positive review|
 |#9033|Singular does not try to build 64-bit on OpenSolaris x64|yes|yes|yes|#9160 does not solve it, but goes some way towards a solution|
@@ -35,7 +35,7 @@
 |#9042|Cython fails to build on OpenSolaris x64|unknown|unknown|yes|Positive review|
 |#9043|lcalc failing to build on OpenSolaris x64.|yes|yes|yes|Positive review|
 |#9040|fatal relocation error installing R on OpenSolaris|unknown|unknown|yes|We many need to build R with SunStudio on OpenSolaris or Solaris 10 on x64|
-|#9097|c_lib in Sage library fails to build on OpenSolaris x64|unknown|unknown|yes|Looks easy - despite it's SCons!|
+|#9097|c_lib in Sage library fails to build on OpenSolaris x64|unknown|unknown|yes|Needs review|
 |#9098|gap buillds 32-bit on OpenSolaris when SAGE64=yes|yes|yes|yes|The -m64 option is not added on any platform. Should be easy to fix.|
 |#9099|Maxima fails to build on OpenSolaris x64, though ECL does.|unknown|unknown|yes||
 |#9100|scipy is probably building part 32-bit on OpenSolaris x64 when SAGE64=yes|unknown|unknown|yes|This is odd - see trac ticket|

bac7d3ea-3f1b-4826-8464-f0b53d5e12d2 commented 14 years ago

Description changed:

--- 
+++ 
@@ -23,7 +23,7 @@
 |#9008|Update zlib to latest upstream, and clean up spkg-install|yes|unknown|yes|**positive review**|
 |#9009|Mercurial is not building 64-bit with OpenSolaris with SAGE64=yes.|yes|yes|yes|Now fixed|
 |#9022|python fails to build _socket on OpenSolaris x64, so ipython fails to build.|unknown|unknown|yes|fixed by #9041 which has positive review|
-|#9023|ghmm is buiding 32-bit on OpenSolaris x64 even when SAGE64 is set to "yes"|yes|yes|yes|needs work|
+|#9023|ghmm needs compiler flag -m64 when SAGE64 is set to "yes"|yes|yes|yes|**needs review**|
 |#9024|tachyon is buiding 32-bit on OpenSolaris x64 even when SAGE64 is set to "yes"|yes|yes|yes|Now fixed|
 |#9025|PALP is building 32-bit on OpenSolaris - probably other platforms too.|yes|yes|yes|Positive review|
 |#9029|sympow is buiding 32-bit on OpenSolaris x64 even when SAGE64 is set to "yes"|yes|yes|yes|Fixed|

bac7d3ea-3f1b-4826-8464-f0b53d5e12d2 commented 14 years ago

Description changed:

--- 
+++ 
@@ -35,8 +35,8 @@
 |#9042|Cython fails to build on OpenSolaris x64|unknown|unknown|yes|Positive review|
 |#9043|lcalc failing to build on OpenSolaris x64.|yes|yes|yes|Positive review|
 |#9040|fatal relocation error installing R on OpenSolaris|unknown|unknown|yes|We many need to build R with SunStudio on OpenSolaris or Solaris 10 on x64|
-|#9097|c_lib in Sage library fails to build on OpenSolaris x64|unknown|unknown|yes|Needs review|
-|#9098|gap buillds 32-bit on OpenSolaris when SAGE64=yes|yes|yes|yes|The -m64 option is not added on any platform. Should be easy to fix.|
+|#9097|c_lib in Sage library fails to build on OpenSolaris x64|unknown|unknown|yes|**Needs review**|
+|#9098|gap buillds 32-bit on OpenSolaris when SAGE64=yes|yes|yes|yes|**Needs review**|
 |#9099|Maxima fails to build on OpenSolaris x64, though ECL does.|unknown|unknown|yes||
 |#9100|scipy is probably building part 32-bit on OpenSolaris x64 when SAGE64=yes|unknown|unknown|yes|This is odd - see trac ticket|
 |#9101|linbox reports "ERROR: BLAS not found!" on Solaris 10 SPARC (64-bit build).|yes|unknown|no|It should use ATLAS|

bac7d3ea-3f1b-4826-8464-f0b53d5e12d2 commented 14 years ago

Description changed:

--- 
+++ 
@@ -36,7 +36,7 @@
 |#9043|lcalc failing to build on OpenSolaris x64.|yes|yes|yes|Positive review|
 |#9040|fatal relocation error installing R on OpenSolaris|unknown|unknown|yes|We many need to build R with SunStudio on OpenSolaris or Solaris 10 on x64|
 |#9097|c_lib in Sage library fails to build on OpenSolaris x64|unknown|unknown|yes|**Needs review**|
-|#9098|gap buillds 32-bit on OpenSolaris when SAGE64=yes|yes|yes|yes|**Needs review**|
+|#9098|gap buillds 32-bit on OpenSolaris when SAGE64=yes|yes|yes|yes|Positive review|
 |#9099|Maxima fails to build on OpenSolaris x64, though ECL does.|unknown|unknown|yes||
 |#9100|scipy is probably building part 32-bit on OpenSolaris x64 when SAGE64=yes|unknown|unknown|yes|This is odd - see trac ticket|
 |#9101|linbox reports "ERROR: BLAS not found!" on Solaris 10 SPARC (64-bit build).|yes|unknown|no|It should use ATLAS|

bac7d3ea-3f1b-4826-8464-f0b53d5e12d2 commented 14 years ago

Description changed:

--- 
+++ 
@@ -19,6 +19,7 @@
 |#7861|pynac not building on Open Solaris x64 (32-bit/64-bit mixup)|unknown|unknown|yes||
 |#7864|libfplll tries to link 64-bit objects to 32-bit libstdc++.so|unknown|unknown|yes||
 |#7982|sage_fortran is not working properly on Open Solaris x64 in 64 bit mode.|yes|yes|yes|**positive review**|
+|#8086|numpy fails to build on Open Solaris x64 - 32 / 64-bit mixup|unknown|unknown|yes|Japp's fix at #8086 does not work for David Kirkby|
 |#8089|ecl 9.10.2-20091105cvs.p1 faiils to build on Open Solaris x64|unknown|unknown|yes|Solved by #8951, which has positive review.|
 |#9008|Update zlib to latest upstream, and clean up spkg-install|yes|unknown|yes|**positive review**|
 |#9009|Mercurial is not building 64-bit with OpenSolaris with SAGE64=yes.|yes|yes|yes|Now fixed|

bac7d3ea-3f1b-4826-8464-f0b53d5e12d2 commented 14 years ago

Description changed:

--- 
+++ 
@@ -40,4 +40,4 @@
 |#9098|gap buillds 32-bit on OpenSolaris when SAGE64=yes|yes|yes|yes|Positive review|
 |#9099|Maxima fails to build on OpenSolaris x64, though ECL does.|unknown|unknown|yes||
 |#9100|scipy is probably building part 32-bit on OpenSolaris x64 when SAGE64=yes|unknown|unknown|yes|This is odd - see trac ticket|
-|#9101|linbox reports "ERROR: BLAS not found!" on Solaris 10 SPARC (64-bit build).|yes|unknown|no|It should use ATLAS|
+|#9101|linbox reports "ERROR: BLAS not found!" on Solaris 10 SPARC (64-bit build).|yes|unknown|no|It should use ATLAS, which mis-compiled as 32-bit|

bac7d3ea-3f1b-4826-8464-f0b53d5e12d2 commented 14 years ago

Description changed:

--- 
+++ 
@@ -41,3 +41,6 @@
 |#9099|Maxima fails to build on OpenSolaris x64, though ECL does.|unknown|unknown|yes||
 |#9100|scipy is probably building part 32-bit on OpenSolaris x64 when SAGE64=yes|unknown|unknown|yes|This is odd - see trac ticket|
 |#9101|linbox reports "ERROR: BLAS not found!" on Solaris 10 SPARC (64-bit build).|yes|unknown|no|It should use ATLAS, which mis-compiled as 32-bit|
+|#9397|Resolve corrupted patches to permit Singular to build on Solaris x86/x64|no|yes|yes|**Needs review**|
+|#9399|Remove Sun-specific junk in rings/finite_rings/stdint.h|no|yes|yes|**Needs review**|
+

jhpalmieri commented 14 years ago

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.

bac7d3ea-3f1b-4826-8464-f0b53d5e12d2 commented 14 years ago

comment:25

Replying to @jhpalmieri:

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.

Yes, that is still true.

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.

I agree, though that is a different issue to this ticket, which is for build problems on 64-bit Solaris. The supported platforms issue needs resolving for Solaris, as it does for Linux, OS X and Windows. I created #9487 to cover this.

Dave

bac7d3ea-3f1b-4826-8464-f0b53d5e12d2 commented 14 years ago

Description changed:

--- 
+++ 
@@ -16,31 +16,31 @@
 |        |         |             |           |                    |         |
 |--------|---------|-------------|-----------|--------------------|---------|
 |**Trac**|**Title**|**S10 SPARC**|**S10 x64**| **OpenSolaris x64**|**Notes**|
-|#7861|pynac not building on Open Solaris x64 (32-bit/64-bit mixup)|unknown|unknown|yes||
-|#7864|libfplll tries to link 64-bit objects to 32-bit libstdc++.so|unknown|unknown|yes||
-|#7982|sage_fortran is not working properly on Open Solaris x64 in 64 bit mode.|yes|yes|yes|**positive review**|
-|#8086|numpy fails to build on Open Solaris x64 - 32 / 64-bit mixup|unknown|unknown|yes|Japp's fix at #8086 does not work for David Kirkby|
-|#8089|ecl 9.10.2-20091105cvs.p1 faiils to build on Open Solaris x64|unknown|unknown|yes|Solved by #8951, which has positive review.|
-|#9008|Update zlib to latest upstream, and clean up spkg-install|yes|unknown|yes|**positive review**|
-|#9009|Mercurial is not building 64-bit with OpenSolaris with SAGE64=yes.|yes|yes|yes|Now fixed|
-|#9022|python fails to build _socket on OpenSolaris x64, so ipython fails to build.|unknown|unknown|yes|fixed by #9041 which has positive review|
-|#9023|ghmm needs compiler flag -m64 when SAGE64 is set to "yes"|yes|yes|yes|**needs review**|
-|#9024|tachyon is buiding 32-bit on OpenSolaris x64 even when SAGE64 is set to "yes"|yes|yes|yes|Now fixed|
-|#9025|PALP is building 32-bit on OpenSolaris - probably other platforms too.|yes|yes|yes|Positive review|
-|#9029|sympow is buiding 32-bit on OpenSolaris x64 even when SAGE64 is set to "yes"|yes|yes|yes|Fixed|
-|#9030|rubiks is building part 32-bit and part 64-bit on OpenSolaris x64.|yes|yes|yes|positive review|
-|#9033|Singular does not try to build 64-bit on OpenSolaris x64|yes|yes|yes|#9160 does not solve it, but goes some way towards a solution|
+|#7861|pynac not building on Open Solaris x64 (32-bit/64-bit mixup)|unknown|unknown|yes|**Needs review**|
+|#7864|libfplll tries to link 64-bit objects to 32-bit libstdc++.so|unknown|unknown|yes|**Needs review**|
+|#7982|sage_fortran is not working properly on Open Solaris x64 in 64 bit mode.|yes|yes|yes||
+|#8086|numpy fails to build on Open Solaris x64 - 32 / 64-bit mixup|unknown|unknown|yes|**Needs review**|
+|#8089|ecl 9.10.2-20091105cvs.p1 fails to build on Open Solaris x64|unknown|unknown|yes||
+|#9008|Update zlib to latest upstream, and clean up spkg-install|yes|unknown|yes||
+|#9009|Mercurial is not building 64-bit with OpenSolaris with SAGE64=yes.|yes|yes|yes||
+|#9022|python fails to build _socket on OpenSolaris x64, so ipython fails to build.|unknown|unknown|yes||
+|#9023|ghmm needs compiler flag -m64 when SAGE64 is set to "yes"|yes|yes|yes||
+|#9024|tachyon is buiding 32-bit on OpenSolaris x64 even when SAGE64 is set to "yes"|yes|yes|yes||
+|#9025|PALP is building 32-bit on OpenSolaris - probably other platforms too.|yes|yes|yes||
+|#9029|sympow is buiding 32-bit on OpenSolaris x64 even when SAGE64 is set to "yes"|yes|yes|yes||
+|#9030|rubiks is building part 32-bit and part 64-bit on OpenSolaris x64.|yes|yes|yes||
+|#9033|Singular does not try to build 64-bit on OpenSolaris x64|yes|yes|yes|Should be fixed by updated Singular at #8059|
 |#9034|flintqs builds as 32-bit despite SAGE64=yes on OpenSolaris x64|yes|yes|yes|Positive review|
 |#9037|pynac fails to build on 64-bit OpenSolaris x64.|yes|yes|yes|Pynac 0.2 now in Sage. That must be fixed - see #8903 too|
-|#9041|python fails to build _socket on OpenSolaris x64, so pygments fails to build.|unknown|unknown|yes|positive review|
-|#9042|Cython fails to build on OpenSolaris x64|unknown|unknown|yes|Positive review|
-|#9043|lcalc failing to build on OpenSolaris x64.|yes|yes|yes|Positive review|
+|#9041|python fails to build _socket on OpenSolaris x64, so pygments fails to build.|unknown|unknown|yes||
+|#9042|Cython fails to build on OpenSolaris x64|unknown|unknown|yes||
+|#9043|lcalc failing to build on OpenSolaris x64.|yes|yes|yes||
 |#9040|fatal relocation error installing R on OpenSolaris|unknown|unknown|yes|We many need to build R with SunStudio on OpenSolaris or Solaris 10 on x64|
-|#9097|c_lib in Sage library fails to build on OpenSolaris x64|unknown|unknown|yes|**Needs review**|
-|#9098|gap buillds 32-bit on OpenSolaris when SAGE64=yes|yes|yes|yes|Positive review|
+|#9097|c_lib in Sage library fails to build on OpenSolaris x64|unknown|unknown|yes||
+|#9098|gap buillds 32-bit on OpenSolaris when SAGE64=yes|yes|yes|yes||
 |#9099|Maxima fails to build on OpenSolaris x64, though ECL does.|unknown|unknown|yes||
 |#9100|scipy is probably building part 32-bit on OpenSolaris x64 when SAGE64=yes|unknown|unknown|yes|This is odd - see trac ticket|
 |#9101|linbox reports "ERROR: BLAS not found!" on Solaris 10 SPARC (64-bit build).|yes|unknown|no|It should use ATLAS, which mis-compiled as 32-bit|
-|#9397|Resolve corrupted patches to permit Singular to build on Solaris x86/x64|no|yes|yes|**Needs review**|
-|#9399|Remove Sun-specific junk in rings/finite_rings/stdint.h|no|yes|yes|**Needs review**|
+|#9397|Resolve corrupted patches to permit Singular to build on Solaris x86/x64|no|yes|yes|Should be fixed by update of Singular at #8059 |
+|#9399|Remove Sun-specific junk in rings/finite_rings/stdint.h|no|yes|yes||

jhpalmieri commented 14 years ago

Description changed:

--- 
+++ 
@@ -43,4 +43,6 @@
 |#9101|linbox reports "ERROR: BLAS not found!" on Solaris 10 SPARC (64-bit build).|yes|unknown|no|It should use ATLAS, which mis-compiled as 32-bit|
 |#9397|Resolve corrupted patches to permit Singular to build on Solaris x86/x64|no|yes|yes|Should be fixed by update of Singular at #8059 |
 |#9399|Remove Sun-specific junk in rings/finite_rings/stdint.h|no|yes|yes||
+|#9358|zn_poly passes all tests on on Solaris 10 64-bit SPARC, but fails to install|no|yes|yes||

+

bac7d3ea-3f1b-4826-8464-f0b53d5e12d2 commented 14 years ago

Description changed:

--- 
+++ 
@@ -41,8 +41,8 @@
 |#9099|Maxima fails to build on OpenSolaris x64, though ECL does.|unknown|unknown|yes||
 |#9100|scipy is probably building part 32-bit on OpenSolaris x64 when SAGE64=yes|unknown|unknown|yes|This is odd - see trac ticket|
 |#9101|linbox reports "ERROR: BLAS not found!" on Solaris 10 SPARC (64-bit build).|yes|unknown|no|It should use ATLAS, which mis-compiled as 32-bit|
-|#9397|Resolve corrupted patches to permit Singular to build on Solaris x86/x64|no|yes|yes|Should be fixed by update of Singular at #8059 |
+|#9358|zn_poly passes all tests on on Solaris 10 64-bit SPARC, but fails to install|no|yes|yes||
+|#9397|Resolve corrupted patches to permit Singular to build on Solaris x86/x64|yes|unknown|no|Should be fixed by update of Singular at #8059 |
 |#9399|Remove Sun-specific junk in rings/finite_rings/stdint.h|no|yes|yes||
-|#9358|zn_poly passes all tests on on Solaris 10 64-bit SPARC, but fails to install|no|yes|yes||

bac7d3ea-3f1b-4826-8464-f0b53d5e12d2 commented 14 years ago

comment:28

John, I see you updated the table to include #9358. Note however note the comments above.

Please put

"yes" if the trac ticket is known, or one might reasonably expect it to cause a problem on a particular platform.
"no" 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.

So the znpoly issue should have a "yes" in the first column. I've also tried to keep them in order of trac number, though that's not a big deal.

Perhaps changing "yes" -> Bug and No->OK might be sensible, as I can see this could be confusing. That would be trivial to change with a sed script. What do you think of that, or can you think of a better, less confusing solution?

Dave

jhpalmieri commented 14 years ago

comment:29

Well, znpoly builds fine on t2, so shouldn't there be a "no" in the first column? It has problems 64-bit on t2, so "yes" in the second column. The third column is the only one I should probably change, to "unknown", since I don't know how znpoly behaves on OpenSolaris. Maybe you know about that, though?

jhpalmieri commented 14 years ago

comment:30

Oh, I see the whole ticket is 64-bit specific. Never mind.

jhpalmieri commented 14 years ago

Description changed:

--- 
+++ 
@@ -41,7 +41,7 @@
 |#9099|Maxima fails to build on OpenSolaris x64, though ECL does.|unknown|unknown|yes||
 |#9100|scipy is probably building part 32-bit on OpenSolaris x64 when SAGE64=yes|unknown|unknown|yes|This is odd - see trac ticket|
 |#9101|linbox reports "ERROR: BLAS not found!" on Solaris 10 SPARC (64-bit build).|yes|unknown|no|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|no|yes|yes||
+|#9358|zn_poly passes all tests on on Solaris 10 64-bit SPARC, but fails to install|yes|yes|unknown||
 |#9397|Resolve corrupted patches to permit Singular to build on Solaris x86/x64|yes|unknown|no|Should be fixed by update of Singular at #8059 |
 |#9399|Remove Sun-specific junk in rings/finite_rings/stdint.h|no|yes|yes||

bac7d3ea-3f1b-4826-8464-f0b53d5e12d2 commented 14 years ago

comment:31

Replying to @jhpalmieri:

Oh, I see the whole ticket is 64-bit specific. Never mind.

We could add a couple of columns, for 32-bit OpenSolaris and 32-bit Solaris 10 x86. I wanted to avoid having. Obviously the title would need changing, but that is no big deal. What do you think to that - put them all on one ticket?

bac7d3ea-3f1b-4826-8464-f0b53d5e12d2 commented 14 years ago

comment:32

Replying to @sagetrac-drkirkby:

Replying to @jhpalmieri:

Oh, I see the whole ticket is 64-bit specific. Never mind.

We could add a couple of columns, for 32-bit OpenSolaris and 32-bit Solaris 10 x86. I wanted to avoid having.

I did not finish that sentence. Just forget it. It seems to me adding a couple of columns might be sensible.

Dave

jhpalmieri commented 14 years ago

Description changed:

--- 
+++ 
@@ -16,10 +16,10 @@
 |        |         |             |           |                    |         |
 |--------|---------|-------------|-----------|--------------------|---------|
 |**Trac**|**Title**|**S10 SPARC**|**S10 x64**| **OpenSolaris x64**|**Notes**|
-|#7861|pynac not building on Open Solaris x64 (32-bit/64-bit mixup)|unknown|unknown|yes|**Needs review**|
-|#7864|libfplll tries to link 64-bit objects to 32-bit libstdc++.so|unknown|unknown|yes|**Needs review**|
+|#7861|pynac not building on Open Solaris x64 (32-bit/64-bit mixup)|unknown|unknown|yes|**Positive review**|
+|#7864|libfplll tries to link 64-bit objects to 32-bit libstdc++.so|unknown|unknown|yes|**Positive review**|
 |#7982|sage_fortran is not working properly on Open Solaris x64 in 64 bit mode.|yes|yes|yes||
-|#8086|numpy fails to build on Open Solaris x64 - 32 / 64-bit mixup|unknown|unknown|yes|**Needs review**|
+|#8086|numpy fails to build on Open Solaris x64 - 32 / 64-bit mixup|unknown|unknown|yes|**Positive review**|
 |#8089|ecl 9.10.2-20091105cvs.p1 fails to build on Open Solaris x64|unknown|unknown|yes||
 |#9008|Update zlib to latest upstream, and clean up spkg-install|yes|unknown|yes||
 |#9009|Mercurial is not building 64-bit with OpenSolaris with SAGE64=yes.|yes|yes|yes||

jhpalmieri commented 14 years ago

Description changed:

--- 
+++ 
@@ -38,7 +38,7 @@
 |#9040|fatal relocation error installing R on OpenSolaris|unknown|unknown|yes|We many need to build R with SunStudio on OpenSolaris or Solaris 10 on x64|
 |#9097|c_lib in Sage library fails to build on OpenSolaris x64|unknown|unknown|yes||
 |#9098|gap buillds 32-bit on OpenSolaris when SAGE64=yes|yes|yes|yes||
-|#9099|Maxima fails to build on OpenSolaris x64, though ECL does.|unknown|unknown|yes||
+|#9099|Maxima fails to build on OpenSolaris x64, though ECL does.|no|yes|yes||
 |#9100|scipy is probably building part 32-bit on OpenSolaris x64 when SAGE64=yes|unknown|unknown|yes|This is odd - see trac ticket|
 |#9101|linbox reports "ERROR: BLAS not found!" on Solaris 10 SPARC (64-bit build).|yes|unknown|no|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|yes|yes|unknown||

bac7d3ea-3f1b-4826-8464-f0b53d5e12d2 commented 14 years ago

Description changed:

--- 
+++ 
@@ -44,5 +44,5 @@
 |#9358|zn_poly passes all tests on on Solaris 10 64-bit SPARC, but fails to install|yes|yes|unknown||
 |#9397|Resolve corrupted patches to permit Singular to build on Solaris x86/x64|yes|unknown|no|Should be fixed by update of Singular at #8059 |
 |#9399|Remove Sun-specific junk in rings/finite_rings/stdint.h|no|yes|yes||
+|#9643|Force ECL to disable assembly code on Solaris 10 x86 as it does on OpenSolaris|no|yes|no||

-

bac7d3ea-3f1b-4826-8464-f0b53d5e12d2 commented 14 years ago

Description changed:

--- 
+++ 
@@ -41,7 +41,7 @@
 |#9099|Maxima fails to build on OpenSolaris x64, though ECL does.|no|yes|yes||
 |#9100|scipy is probably building part 32-bit on OpenSolaris x64 when SAGE64=yes|unknown|unknown|yes|This is odd - see trac ticket|
 |#9101|linbox reports "ERROR: BLAS not found!" on Solaris 10 SPARC (64-bit build).|yes|unknown|no|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|yes|yes|unknown||
+|#9358|zn_poly passes all tests on on Solaris 10 64-bit SPARC, but fails to install|yes|yes|yes|**needs review**|
 |#9397|Resolve corrupted patches to permit Singular to build on Solaris x86/x64|yes|unknown|no|Should be fixed by update of Singular at #8059 |
 |#9399|Remove Sun-specific junk in rings/finite_rings/stdint.h|no|yes|yes||
 |#9643|Force ECL to disable assembly code on Solaris 10 x86 as it does on OpenSolaris|no|yes|no||

bac7d3ea-3f1b-4826-8464-f0b53d5e12d2 commented 14 years ago

Description changed:

--- 
+++ 
@@ -1,4 +1,4 @@
-The following is list of the issues that are currently preventing Sage build on either
+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')
@@ -7,8 +7,8 @@
 The list can be added as new problems are found, making what William calls a 'metaticket' 

 Please put 
-* "yes" if the trac ticket is known, or one might reasonably expect it to cause a problem on a particular platform.
-* "no" if one knows the problem does not exist on that platform, or one reasonably expect it will not be an issue. 
+* "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)
@@ -16,33 +16,34 @@
 |        |         |             |           |                    |         |
 |--------|---------|-------------|-----------|--------------------|---------|
 |**Trac**|**Title**|**S10 SPARC**|**S10 x64**| **OpenSolaris x64**|**Notes**|
-|#7861|pynac not building on Open Solaris x64 (32-bit/64-bit mixup)|unknown|unknown|yes|**Positive review**|
-|#7864|libfplll tries to link 64-bit objects to 32-bit libstdc++.so|unknown|unknown|yes|**Positive review**|
-|#7982|sage_fortran is not working properly on Open Solaris x64 in 64 bit mode.|yes|yes|yes||
-|#8086|numpy fails to build on Open Solaris x64 - 32 / 64-bit mixup|unknown|unknown|yes|**Positive review**|
-|#8089|ecl 9.10.2-20091105cvs.p1 fails to build on Open Solaris x64|unknown|unknown|yes||
-|#9008|Update zlib to latest upstream, and clean up spkg-install|yes|unknown|yes||
-|#9009|Mercurial is not building 64-bit with OpenSolaris with SAGE64=yes.|yes|yes|yes||
-|#9022|python fails to build _socket on OpenSolaris x64, so ipython fails to build.|unknown|unknown|yes||
-|#9023|ghmm needs compiler flag -m64 when SAGE64 is set to "yes"|yes|yes|yes||
-|#9024|tachyon is buiding 32-bit on OpenSolaris x64 even when SAGE64 is set to "yes"|yes|yes|yes||
-|#9025|PALP is building 32-bit on OpenSolaris - probably other platforms too.|yes|yes|yes||
-|#9029|sympow is buiding 32-bit on OpenSolaris x64 even when SAGE64 is set to "yes"|yes|yes|yes||
-|#9030|rubiks is building part 32-bit and part 64-bit on OpenSolaris x64.|yes|yes|yes||
-|#9033|Singular does not try to build 64-bit on OpenSolaris x64|yes|yes|yes|Should be fixed by updated Singular at #8059|
-|#9034|flintqs builds as 32-bit despite SAGE64=yes on OpenSolaris x64|yes|yes|yes|Positive review|
-|#9037|pynac fails to build on 64-bit OpenSolaris x64.|yes|yes|yes|Pynac 0.2 now in Sage. That must be fixed - see #8903 too|
-|#9041|python fails to build _socket on OpenSolaris x64, so pygments fails to build.|unknown|unknown|yes||
-|#9042|Cython fails to build on OpenSolaris x64|unknown|unknown|yes||
-|#9043|lcalc failing to build on OpenSolaris x64.|yes|yes|yes||
-|#9040|fatal relocation error installing R on OpenSolaris|unknown|unknown|yes|We many need to build R with SunStudio on OpenSolaris or Solaris 10 on x64|
-|#9097|c_lib in Sage library fails to build on OpenSolaris x64|unknown|unknown|yes||
-|#9098|gap buillds 32-bit on OpenSolaris when SAGE64=yes|yes|yes|yes||
-|#9099|Maxima fails to build on OpenSolaris x64, though ECL does.|no|yes|yes||
-|#9100|scipy is probably building part 32-bit on OpenSolaris x64 when SAGE64=yes|unknown|unknown|yes|This is odd - see trac ticket|
-|#9101|linbox reports "ERROR: BLAS not found!" on Solaris 10 SPARC (64-bit build).|yes|unknown|no|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|yes|yes|yes|**needs review**|
-|#9397|Resolve corrupted patches to permit Singular to build on Solaris x86/x64|yes|unknown|no|Should be fixed by update of Singular at #8059 |
-|#9399|Remove Sun-specific junk in rings/finite_rings/stdint.h|no|yes|yes||
-|#9643|Force ECL to disable assembly code on Solaris 10 x86 as it does on OpenSolaris|no|yes|no||
+|#7861|pynac not building on Open Solaris x64 (32-bit/64-bit mixup)|unknown|unknown|bug|**Positive review**|
+|#7864|libfplll tries to link 64-bit objects to 32-bit libstdc++.so|unknown|unknown|bug|**Positive review**|
+|#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|**Positive review**|
+|#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||
+|#9033|Singular does not try to build 64-bit on OpenSolaris x64|bug|bug|bug|Should be fixed by updated Singular at #8059|
+|#9034|flintqs builds as 32-bit despite SAGE64=yes on OpenSolaris x64|bug|bug|bug|Positive review|
+|#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|**needs 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||
+|#9643|Force ECL to disable assembly code on Solaris 10 x86 as it does on OpenSolaris|OK|bug|OK|OpenSolaris resolved in #9474|

+

bac7d3ea-3f1b-4826-8464-f0b53d5e12d2 commented 14 years ago

Description changed:

--- 
+++ 
@@ -44,6 +44,7 @@
 |#9358|zn_poly passes all tests on on Solaris 10 64-bit SPARC, but fails to install|bug|bug|bug|**needs 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|OpenSolaris resolved in #9474|

bac7d3ea-3f1b-4826-8464-f0b53d5e12d2 commented 14 years ago

Description changed:

--- 
+++ 
@@ -29,7 +29,6 @@
 |#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||
-|#9033|Singular does not try to build 64-bit on OpenSolaris x64|bug|bug|bug|Should be fixed by updated Singular at #8059|
 |#9034|flintqs builds as 32-bit despite SAGE64=yes on OpenSolaris x64|bug|bug|bug|Positive review|
 |#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|

jhpalmieri commented 14 years ago

Description changed:

--- 
+++ 
@@ -40,7 +40,7 @@
 |#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|**needs review**|
+|#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||

bac7d3ea-3f1b-4826-8464-f0b53d5e12d2 commented 14 years ago

Description changed:

--- 
+++ 
@@ -16,10 +16,10 @@
 |        |         |             |           |                    |         |
 |--------|---------|-------------|-----------|--------------------|---------|
 |**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|**Positive review**|
-|#7864|libfplll tries to link 64-bit objects to 32-bit libstdc++.so|unknown|unknown|bug|**Positive review**|
+|#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|**Positive review**|
+|#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||
@@ -29,7 +29,7 @@
 |#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|Positive review|
+|#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||
@@ -44,6 +44,6 @@
 |#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|OpenSolaris resolved in #9474|
+|#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)|

mkoeppe commented 4 years ago

comment:46

Should be closed as outdated.

dimpase commented 4 years ago

Reviewer: Dima Pasechnik

fchapoton commented 4 years ago

comment:48

agreed