New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Update eclib to latest upstream release #10993
Comments
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
comment:5
Working on this at Sage Days 35 (18 Dec 2011) -- will be ready for review soon. |
comment:6
Ready for review. Thanks to Simon King and Paul Zimmermann for help adding the ldconfig line to the spkg-install script, which was necessary after Giovanni Mascellani mascellani@poisson.phc.unipi.it added the SONAME lines to my Makefiles. |
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
comment:9
I tested with sage-4.8.alpha4. The patche applies fine, |
Reviewer: fschulze |
Work Issues: rebase |
comment:10
I'm afraid this needs to be rebased to #11354. |
Dependencies: #11354 |
comment:134
Replying to @nexttime:
Like this: #include <cstdlib>
#include <iostream>
using std::getenv;
const char *getenv(const char *var_name, const char *default_value)
{
const char *val = getenv(var_name);
return val ? val : default_value;
}
int main(void)
{
std::cout << "getenv(\"FOO\"): " << getenv("FOO") << std::endl;
std::cout.clear();
std::cout << std::endl;
std::cout << "getenv(\"FOO\", \"bar\"): " << getenv("FOO", "bar") << std::endl;
return 0;
} which yields
|
comment:135
Replying to @JohnCremona:
Just waiting for tests (actually of the April 19th spkg) on Solaris to finish... (They passed on Linux x86_64.) |
comment:136
The link in the description was still referencing the 20120419 spkg. |
This comment has been minimized.
This comment has been minimized.
comment:137
Replying to @nexttime:
|
comment:138
Replying to @nexttime:
Don't know whether that's a regression, but some tests take really long (as mentioned), at least on mark2:
|
comment:139
Replying to @nexttime:
beta13 + MPIR 2.4.0 + MPFR 3.1.0 with all upstream patches; all (relevant) packages compiled with GCC 4.7.0 with |
comment:140
None of the changes to eclib would make a noticeable difference in the timings of any Sage doctests. Many other changes in this testing version of Sage do have such an effect. Unless you can prove me wrong ( by giving the timings on exactly the same Sage version without the new eclib) ... |
comment:141
Replying to @JohnCremona:
Nope. I see a slight slowdown for Anyway, someone else would have to prove you wrong.., ;-) |
comment:142
Ok, latest spkg (20120421) looks clean, and also passes all relevant tests on Solaris SPARC (32-bit) as well as on Linux x86_64, so finally setting it to positive review, assuming Volker agrees. (I guess John also prefers to [hopefully] get it into Sage 5.0 as is, rather than adding even more upstream changes which would again need review and testing.) |
Changed reviewer from Frithjof Schulze, Jeroen Demeyer to Frithjof Schulze, Jeroen Demeyer, Volker Braun, Leif Leonhardy |
comment:143
Sounds great! |
comment:144
Replying to @vbraun:
|
comment:145
Just in case you're still listening, the new spkg also passes all [IMHO] relevant tests on Linux ppc64 (SLES 11, POWER7) with Sage 5.0.beta9 and #12829, GCC 4.6.3. (And I haven't observed speed regressions; this time I've run the tests [in parallel] with the old and the new spkg with the same Sage installation, compiler and flags.) |
Changed work issues from ldconfig in spkg-install to none |
comment:148
I checked that eclib-20120421.spkg compiles well on my ARM system running ubuntu. The patch applied with fuzz, and "sage -ba-force" was a success too. |
comment:149
Replying to @vbraun:
|
This comment has been minimized.
This comment has been minimized.
comment:150
Yet another upstream update? Note that the link target in the description is still |
comment:151
It's a tiny bug fix in something not used in Sage, and as the target has moved to 5.1 I thought I would change the spkg. Sorry if that causes more work in testing. I'm finding it very hard to edit trac tickets since the version was changed, especially on a netbook while travelling. I only see one place on the ticket where the spkg name is given, in the Description, and I changed that. If there's another one somewhere (why would there be two?) and you can change it to match, please do. |
This comment has been minimized.
This comment has been minimized.
comment:152
Replying to @JohnCremona:
No, but in the description we currently again have a link target different to its name:
It's sufficient to just use
(where the brackets are optional if you don't specify an [alternate] name for the link, but I always use them). Made that change, although technically the new spkg would need new review. |
Merged: sage-5.1.beta0 |
eclib has changed a lot since the last upgrade, mostly in parts which do not affect Sage much, but there are exceptions, for example the handling of bounds in saturation for elliptic curves over Q (see #10840).
As well as various changes in the source code, the new distribution in the new spkg now uses autotools throughout which should make building on different systems easier.
After installing it (and before "sage -b") apply the patch also.
spkg: http://homepages.warwick.ac.uk/staff/J.E.Cremona/ftp/progs/eclib-20120428.spkg
Apply: attachment: trac_10993-eclib.patch
Depends on #11354
CC: @kiwifb @sagetrac-fschulze
Component: packages: standard
Author: John Cremona
Reviewer: Frithjof Schulze, Jeroen Demeyer, Volker Braun, Leif Leonhardy
Merged: sage-5.1.beta0
Issue created by migration from https://trac.sagemath.org/ticket/10993
The text was updated successfully, but these errors were encountered: