Upgrade: cypari2 2.1.2 #31029

videlec · 2020-12-08T09:55:52Z

This is to upgrade cypari to 2.1.2 (which is compatible with pari 2.13 that is about to be upgraded too, see #30801)

https://files.pythonhosted.org/packages/22/d9/4374baf5749257362f6d163096f9f5a3fa2a6f231edbe0d61c7bdf281d6e/cypari2-2.1.2.tar.gz

CC: @dimpase @orlitzky @mkoeppe @slel @antonio-rojas @kiwifb @videlec @tobihan @seblabbe @fchapoton @jhpalmieri

Component: packages: standard

Keywords: sd111

Author: Vincent Delecroix

Branch/Commit: b9aadfd

Reviewer: Matthias Koeppe, John Palmieri

Issue created by migration from https://trac.sagemath.org/ticket/31029

The text was updated successfully, but these errors were encountered:

videlec · 2020-12-08T09:58:30Z

New commits:

`0b18fd0`	`31029: upgrade cypari2 to 2.1.2`
`b9aadfd`	`31029: adapt sage library code to cypari2 version 2.1.2`

videlec · 2020-12-08T09:58:30Z

Commit: b9aadfd

videlec · 2020-12-08T09:58:30Z

Branch: u/vdelecroix/31029

slel · 2020-12-08T11:52:40Z

comment:2

Pushed a tag to launch "GitHub Actions" CI:

mkoeppe · 2020-12-11T03:30:12Z

Reviewer: Matthias Koeppe, ...

mkoeppe · 2020-12-11T03:30:12Z

comment:4

The builds on GH Actions look fine.

Someone else should look at the changes in the ticket -- I don't know anything about PARI.

mkoeppe · 2020-12-11T03:30:22Z

Changed keywords from none to sd111

jhpalmieri · 2020-12-28T20:49:18Z

comment:9

Unfortunately this does not fix my problem with OS X 10.15.7 and Python 3 installed via Homebrew:

  cypari2/gen.c:25980:70: error: too many arguments to function call, expected 3, have 4
    __pyx_v__ret = asympnum0(__pyx_v__expr, __pyx_v_k, __pyx_v__alpha, __pyx_v_precision);
                   ~~~~~~~~~                                           ^~~~~~~~~~~~~~~~~
  /usr/local/include/pari/paripriv.h:101:1: note: 'asympnum0' declared here
  GEN  asympnum0(GEN u, GEN alpha, long prec);
  ^
  cypari2/gen.c:30261:18: error: implicit declaration of function 'bnfcompress' is invalid in C99 [-Werror,-Wimplicit-function-declaration]
    __pyx_v__ret = bnfcompress(__pyx_v__bnf);
                   ^
  cypari2/gen.c:30261:16: warning: incompatible integer to pointer conversion assigning to 'GEN' (aka 'long *') from 'int' [-Wint-conversion]
    __pyx_v__ret = bnfcompress(__pyx_v__bnf);
                 ^ ~~~~~~~~~~~~~~~~~~~~~~~~~
  cypari2/gen.c:84064:52: error: too few arguments to function call, expected 3, have 2
    __pyx_v__ret = lfuntwist(__pyx_v__L, __pyx_v__chi);
                   ~~~~~~~~~                         ^
  /usr/local/include/pari/paridecl.h:4164:1: note: 'lfuntwist' declared here
  GEN     lfuntwist(GEN ldata1, GEN ldata2, long bitprec);
  ^
  cypari2/gen.c:85034:70: error: too many arguments to function call, expected 3, have 4
    __pyx_v__ret = limitnum0(__pyx_v__expr, __pyx_v_k, __pyx_v__alpha, __pyx_v_precision);
                   ~~~~~~~~~                                           ^~~~~~~~~~~~~~~~~
  /usr/local/include/pari/paripriv.h:125:1: note: 'limitnum0' declared here
  GEN  limitnum0(GEN u, GEN alpha, long prec);
  ^
  cypari2/gen.c:108195:18: error: implicit declaration of function 'nfbasis_gp' is invalid in C99 [-Werror,-Wimplicit-function-declaration]
    __pyx_v__ret = nfbasis_gp(__pyx_v__T);
                   ^
  cypari2/gen.c:108195:18: note: did you mean 'nfbasis'?
  /usr/local/include/pari/paridecl.h:2231:9: note: 'nfbasis' declared here
  GEN     nfbasis(GEN x, GEN *y);
          ^
  cypari2/gen.c:108195:16: warning: incompatible integer to pointer conversion assigning to 'GEN' (aka 'long *') from 'int' [-Wint-conversion]
    __pyx_v__ret = nfbasis_gp(__pyx_v__T);
                 ^ ~~~~~~~~~~~~~~~~~~~~~~
  cypari2/gen.c:119493:16: warning: incompatible pointer to integer conversion assigning to 'long' from 'GEN' (aka 'long *'); dereference with * [-Wint-conversion]
    __pyx_v__ret = permorder(__pyx_v__x);
                 ^ ~~~~~~~~~~~~~~~~~~~~~
                   *
  cypari2/gen.c:123101:18: error: implicit declaration of function 'isirreducible' is invalid in C99 [-Werror,-Wimplicit-function-declaration]
    __pyx_v__ret = isirreducible(__pyx_v__pol);
                   ^
  cypari2/gen.c:123101:18: note: did you mean 'polisirreducible'?
  /usr/local/include/pari/paridecl.h:4812:9: note: 'polisirreducible' declared here
  long    polisirreducible(GEN x);
          ^
  cypari2/gen.c:131353:49: error: too few arguments to function call, expected 3, have 2
    __pyx_v__ret = qfbsolve(__pyx_v__Q, __pyx_v__p);
                   ~~~~~~~~                       ^
  /usr/local/include/pari/paridecl.h:1880:1: note: 'qfbsolve' declared here
  GEN     qfbsolve(GEN Q, GEN n, long flag);
  ^
  cypari2/gen.c:141271:69: error: too many arguments to function call, expected 3, have 4
    __pyx_v__ret = rnfkummer(__pyx_v__bnr, __pyx_v__subgp, __pyx_v_d, __pyx_v_precision);
                   ~~~~~~~~~                                          ^~~~~~~~~~~~~~~~~
  /usr/local/include/pari/paridecl.h:4110:1: note: 'rnfkummer' declared here
  GEN     rnfkummer(GEN bnr, GEN subgroup, long prec);
  ^
  cypari2/gen.c:153671:18: error: implicit declaration of function 'zetamult0' is invalid in C99 [-Werror,-Wimplicit-function-declaration]
    __pyx_v__ret = zetamult0(__pyx_v__s, __pyx_v__T, __pyx_v_precision);
                   ^
  cypari2/gen.c:153671:18: note: did you mean 'zetamult'?
  /usr/local/include/pari/paridecl.h:5345:9: note: 'zetamult' declared here
  GEN     zetamult(GEN avec, long prec);
          ^
  cypari2/gen.c:153671:16: warning: incompatible integer to pointer conversion assigning to 'GEN' (aka 'long *') from 'int' [-Wint-conversion]
    __pyx_v__ret = zetamult0(__pyx_v__s, __pyx_v__T, __pyx_v_precision);
                 ^ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
  4 warnings and 9 errors generated.

jhpalmieri · 2020-12-28T21:24:03Z

comment:10

Just for fun I removed all of my homebrew installations and reinstalled a subset of them (brew install aspell autoconf automake boost ecm freetype gcc gd ghostscript glpk gmp gpatch gsl jpeg libmpc libpng libtiff libtool mpfi mpfr nauty openblas pari perl pkg-config ppl r suite-sparse yasm zeromq). Same error.

mkoeppe · 2020-12-29T05:21:30Z

comment:11

And without pari from homebrew?

jhpalmieri · 2020-12-29T06:42:58Z

comment:12

With or without pari from homebrew, Sage builds its own pari, but if I do brew uninstall pari, then cypari builds successfully. Thank you, that's helpful. Can we prevent the build process from picking up the wrong pari somehow?

mkoeppe · 2020-12-29T19:59:42Z

comment:13

Replying to @jhpalmieri:

With or without pari from homebrew, Sage builds its own pari, but if I do brew uninstall pari, then cypari builds successfully. Thank you, that's helpful. Can we prevent the build process from picking up the wrong pari somehow?

This is now #31132.

dimpase · 2020-12-30T00:45:27Z

comment:14

How about biting the bullet and adding the missing Pari parts (i.e. Pari packages) to Homebrew?

jhpalmieri · 2020-12-30T00:57:59Z

comment:15

Unless I'm misunderstanding, that might fix the symptom but not the actual problem. If we go that direction and we don't address the problem of /usr/local leaking into the build, then we have to detect whether all of the relevant pieces of pari are installed in homebrew or none of them are, and refuse to continue if the system is some intermediate state (like mine with pari installed but none of the as yet non-existent packages).

dimpase · 2020-12-31T00:34:17Z

comment:16

Replying to @jhpalmieri:

Unless I'm misunderstanding, that might fix the symptom but not the actual problem. If we go that direction and we don't address the problem of /usr/local leaking into the build, then we have to detect whether all of the relevant pieces of pari are installed in homebrew or none of them are, and refuse to continue if the system is some intermediate state (like mine with pari installed but none of the as yet non-existent packages).

you're right, that's what can be done. (Realistically though, once it's in place, people would just install the whole bunch - one can also have a Sage-components formula to install as many deps as there are)

mkoeppe · 2021-01-06T19:54:03Z

comment:17

Contributing to homebrew packaging is beyond the scope of this ticket.

Still needs review

jhpalmieri · 2021-01-08T00:17:49Z

comment:18

Why remove the doctest

sage: b[0][0].precision()    # in words

?

videlec · 2021-01-08T09:05:22Z

comment:19

Replying to @jhpalmieri:

Why remove the doctest
sage: b[0][0].precision()    # in words
?

sagemath/cypari2@8bd7ca3

jhpalmieri · 2021-01-08T17:34:21Z

comment:20

Great. Let's merge this.

jhpalmieri · 2021-01-08T17:34:21Z

Changed reviewer from Matthias Koeppe, ... to Matthias Koeppe, John Palmieri

vbraun · 2021-01-17T13:46:17Z

Changed branch from u/vdelecroix/31029 to b9aadfd

videlec added this to the sage-9.3 milestone Dec 8, 2020

videlec added c: packages: standard labels Dec 8, 2020

videlec added the s: needs review label Dec 8, 2020

mkoeppe added p: critical / 2 and removed p: major / 3 labels Dec 27, 2020

jhpalmieri added s: positive review and removed s: needs review labels Jan 8, 2021

vbraun removed the s: positive review label Jan 17, 2021

vbraun closed this as completed in 794d068 Jan 17, 2021

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Upgrade: cypari2 2.1.2 #31029

Upgrade: cypari2 2.1.2 #31029

videlec commented Dec 8, 2020

videlec commented Dec 8, 2020

videlec commented Dec 8, 2020

videlec commented Dec 8, 2020

slel commented Dec 8, 2020

mkoeppe commented Dec 11, 2020

mkoeppe commented Dec 11, 2020

mkoeppe commented Dec 11, 2020

jhpalmieri commented Dec 28, 2020

jhpalmieri commented Dec 28, 2020

mkoeppe commented Dec 29, 2020

jhpalmieri commented Dec 29, 2020

mkoeppe commented Dec 29, 2020

dimpase commented Dec 30, 2020

jhpalmieri commented Dec 30, 2020

dimpase commented Dec 31, 2020

mkoeppe commented Jan 6, 2021

jhpalmieri commented Jan 8, 2021

videlec commented Jan 8, 2021

jhpalmieri commented Jan 8, 2021

jhpalmieri commented Jan 8, 2021

vbraun commented Jan 17, 2021

Upgrade: cypari2 2.1.2 #31029

Upgrade: cypari2 2.1.2 #31029

Comments

videlec commented Dec 8, 2020

videlec commented Dec 8, 2020

videlec commented Dec 8, 2020

videlec commented Dec 8, 2020

slel commented Dec 8, 2020

mkoeppe commented Dec 11, 2020

mkoeppe commented Dec 11, 2020

mkoeppe commented Dec 11, 2020

jhpalmieri commented Dec 28, 2020

jhpalmieri commented Dec 28, 2020

mkoeppe commented Dec 29, 2020

jhpalmieri commented Dec 29, 2020

mkoeppe commented Dec 29, 2020

dimpase commented Dec 30, 2020

jhpalmieri commented Dec 30, 2020

dimpase commented Dec 31, 2020

mkoeppe commented Jan 6, 2021

jhpalmieri commented Jan 8, 2021

videlec commented Jan 8, 2021

jhpalmieri commented Jan 8, 2021

jhpalmieri commented Jan 8, 2021

vbraun commented Jan 17, 2021