PARI LLL speed depends on PARI stack size

[jdemeyer]

From a doctest in qqbar.py:

sage: x = polygen(AA)
sage: %time QQbar.polynomial_root(x^5 - x - 1, CIF(RIF(0.1, 0.2), RIF(1.0, 1.1))).imag().exactify()
CPU times: user 28.33 s, sys: 0.02 s, total: 28.35 s
Wall time: 28.38 s
sage: pari.allocatemem(1<<30)
PARI stack size set to 1073741824 bytes
sage: %time QQbar.polynomial_root(x^5 - x - 1, CIF(RIF(0.1, 0.2), RIF(1.0, 1.1))).imag().exactify()
CPU times: user 2.70 s, sys: 0.15 s, total: 2.85 s
Wall time: 2.84 s

Upstream: http://pari.math.u-bordeaux.fr/cgi-bin/bugreport.cgi?bug=1653

Depends on #15653
Depends on #15654

Upstream: Fixed upstream, but not in a stable release.

Component: performance

Branch/Commit: u/rws/ticket/15660 @ d484242

Reviewer: Jeroen Demeyer

Issue created by migration from https://trac.sagemath.org/ticket/15660

[jdemeyer]

Author: Jeroen Demeyer

[jdemeyer]

Dependencies: #15653, #15654

[jdemeyer]

Branch: u/jdemeyer/ticket/15660

[sagetrac-git]

Branch pushed to git repo; I updated commit sha1. New commits:

8c58208

"long time" fixes

[sagetrac-git]

Commit: 8c58208

[jdemeyer]

New commits:

7537427	Fix conversion of zero p-adics to PARI
a955e45	Speed up PARI determinants when stacksize is small
b29b0b7	Speed up PARI LLL when stacksize is small
8c58208	"long time" fixes

[rwst]

Work Issues: rebase

[rwst]

Changed branch from u/jdemeyer/ticket/15660 to u/rws/ticket/15660

[rwst]

Changed commit from 8c58208 to d484242

[rwst]

New commits:

d484242

Merge branch 'develop' into t/15660/ticket/15660

[rwst]

Changed work issues from rebase to none

[jdemeyer]

Reviewer: Jeroen Demeyer

[jdemeyer]

comment:14

Fixed by #16997.

[jdemeyer]

Upstream: Fixed upstream, but not in a stable release.

[jdemeyer]

Changed author from Jeroen Demeyer to none