public inbox for gentoo-science@lists.gentoo.org
 help / color / mirror / Atom feed
* [gentoo-science] blas selection in lapack-reference
@ 2014-01-22  0:52 François Bissey
  2014-01-22  1:47 ` Steven Trogdon
  0 siblings, 1 reply; 5+ messages in thread
From: François Bissey @ 2014-01-22  0:52 UTC (permalink / raw
  To: gentoo-science

We have yet another instance of someone emerging lapack-reference 
without
a valid blas configuration in 
https://bugs.gentoo.org/show_bug.cgi?id=498490
(not their original problem).
I think we should do something about this in lapack-reference and 
possibly
other ebuilds.
In pkg_setup, we could run
eselect blas update
to make sure that a valid configuration is active
and display an informational message about the blas provider
used with the output of
eselect blas show

Who thinks is a good/bad idea?

Francois


^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [gentoo-science] blas selection in lapack-reference
  2014-01-22  0:52 [gentoo-science] blas selection in lapack-reference François Bissey
@ 2014-01-22  1:47 ` Steven Trogdon
  2014-01-22  9:47   ` Reinis Danne
  0 siblings, 1 reply; 5+ messages in thread
From: Steven Trogdon @ 2014-01-22  1:47 UTC (permalink / raw
  To: gentoo-science

On Wed, 22 Jan 2014 13:52:57 +1300
François Bissey <fbissey@slingshot.co.nz> wrote:

> We have yet another instance of someone emerging lapack-reference 
> without
> a valid blas configuration in 
> https://bugs.gentoo.org/show_bug.cgi?id=498490
> (not their original problem).
> I think we should do something about this in lapack-reference and 
> possibly
> other ebuilds.
> In pkg_setup, we could run
> eselect blas update
> to make sure that a valid configuration is active
> and display an informational message about the blas provider
> used with the output of
> eselect blas show
> 
> Who thinks is a good/bad idea?
> 
> Francois
> 

For one, I'm in favor of something like this - though I probably don't have
much say in the matter. I've been bitten by not eselecting {blas, cblas,
lapack} after an upgrade. And I've even posted on this forum the problem.
 
http://article.gmane.org/gmane.linux.gentoo.science/1915
http://article.gmane.org/gmane.linux.gentoo.science/1916

I resolved things, quite accidently, by re-eselecting the important
components which were eselected before the upgrade.

Steve


^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [gentoo-science] blas selection in lapack-reference
  2014-01-22  1:47 ` Steven Trogdon
@ 2014-01-22  9:47   ` Reinis Danne
  2014-01-22  9:54     ` François Bissey
  0 siblings, 1 reply; 5+ messages in thread
From: Reinis Danne @ 2014-01-22  9:47 UTC (permalink / raw
  To: gentoo-science

On Tue, Jan 21, 2014 at 07:47:48PM -0600, Steven Trogdon wrote:
> On Wed, 22 Jan 2014 13:52:57 +1300
> François Bissey <fbissey@slingshot.co.nz> wrote:
> 
> > We have yet another instance of someone emerging lapack-reference 
> > without
> > a valid blas configuration in 
> > https://bugs.gentoo.org/show_bug.cgi?id=498490
> > (not their original problem).
> > I think we should do something about this in lapack-reference and 
> > possibly
> > other ebuilds.
> > In pkg_setup, we could run
> > eselect blas update
> > to make sure that a valid configuration is active
> > and display an informational message about the blas provider
> > used with the output of
> > eselect blas show
> > 
> > Who thinks is a good/bad idea?
> > 
> > Francois
> > 
> 
> For one, I'm in favor of something like this - though I probably don't have
> much say in the matter. I've been bitten by not eselecting {blas, cblas,
> lapack} after an upgrade. And I've even posted on this forum the problem.
>  
> http://article.gmane.org/gmane.linux.gentoo.science/1915
> http://article.gmane.org/gmane.linux.gentoo.science/1916
> 
> I resolved things, quite accidently, by re-eselecting the important
> components which were eselected before the upgrade.
> 
> Steve
> 

I fixed provider selection during upgrades in overlay
(yesterday). Now it should have valid provider set at any time.
With that do you still think it would need this extra eselect
update in ebuild?


Reinis


^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [gentoo-science] blas selection in lapack-reference
  2014-01-22  9:47   ` Reinis Danne
@ 2014-01-22  9:54     ` François Bissey
  2014-01-22 10:02       ` Reinis Danne
  0 siblings, 1 reply; 5+ messages in thread
From: François Bissey @ 2014-01-22  9:54 UTC (permalink / raw
  To: gentoo-science

On 2014-01-22 22:47, Reinis Danne wrote:
> On Tue, Jan 21, 2014 at 07:47:48PM -0600, Steven Trogdon wrote:
>> On Wed, 22 Jan 2014 13:52:57 +1300
>> François Bissey <fbissey@slingshot.co.nz> wrote:
>> 
>> > We have yet another instance of someone emerging lapack-reference
>> > without
>> > a valid blas configuration in
>> > https://bugs.gentoo.org/show_bug.cgi?id=498490
>> > (not their original problem).
>> > I think we should do something about this in lapack-reference and
>> > possibly
>> > other ebuilds.
>> > In pkg_setup, we could run
>> > eselect blas update
>> > to make sure that a valid configuration is active
>> > and display an informational message about the blas provider
>> > used with the output of
>> > eselect blas show
>> >
>> > Who thinks is a good/bad idea?
>> >
>> > Francois
>> >
>> 
>> For one, I'm in favor of something like this - though I probably don't 
>> have
>> much say in the matter. I've been bitten by not eselecting {blas, 
>> cblas,
>> lapack} after an upgrade. And I've even posted on this forum the 
>> problem.
>> 
>> http://article.gmane.org/gmane.linux.gentoo.science/1915
>> http://article.gmane.org/gmane.linux.gentoo.science/1916
>> 
>> I resolved things, quite accidently, by re-eselecting the important
>> components which were eselected before the upgrade.
>> 
>> Steve
>> 
> 
> I fixed provider selection during upgrades in overlay
> (yesterday). Now it should have valid provider set at any time.
> With that do you still think it would need this extra eselect
> update in ebuild?
> 
> 
It seems that the difficult point is when you go from no provider 
available
to one provider available. If that case is fixed then: No

Francois


^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [gentoo-science] blas selection in lapack-reference
  2014-01-22  9:54     ` François Bissey
@ 2014-01-22 10:02       ` Reinis Danne
  0 siblings, 0 replies; 5+ messages in thread
From: Reinis Danne @ 2014-01-22 10:02 UTC (permalink / raw
  To: gentoo-science

On Wed, Jan 22, 2014 at 10:54:48PM +1300, François Bissey wrote:
> On 2014-01-22 22:47, Reinis Danne wrote:
> > On Tue, Jan 21, 2014 at 07:47:48PM -0600, Steven Trogdon wrote:
> >> On Wed, 22 Jan 2014 13:52:57 +1300
> >> François Bissey <fbissey@slingshot.co.nz> wrote:
> >> 
> >> > We have yet another instance of someone emerging lapack-reference
> >> > without
> >> > a valid blas configuration in
> >> > https://bugs.gentoo.org/show_bug.cgi?id=498490
> >> > (not their original problem).
> >> > I think we should do something about this in lapack-reference and
> >> > possibly
> >> > other ebuilds.
> >> > In pkg_setup, we could run
> >> > eselect blas update
> >> > to make sure that a valid configuration is active
> >> > and display an informational message about the blas provider
> >> > used with the output of
> >> > eselect blas show
> >> >
> >> > Who thinks is a good/bad idea?
> >> >
> >> > Francois
> >> >
> >> 
> >> For one, I'm in favor of something like this - though I probably don't 
> >> have
> >> much say in the matter. I've been bitten by not eselecting {blas, 
> >> cblas,
> >> lapack} after an upgrade. And I've even posted on this forum the 
> >> problem.
> >> 
> >> http://article.gmane.org/gmane.linux.gentoo.science/1915
> >> http://article.gmane.org/gmane.linux.gentoo.science/1916
> >> 
> >> I resolved things, quite accidently, by re-eselecting the important
> >> components which were eselected before the upgrade.
> >> 
> >> Steve
> >> 
> > 
> > I fixed provider selection during upgrades in overlay
> > (yesterday). Now it should have valid provider set at any time.
> > With that do you still think it would need this extra eselect
> > update in ebuild?
> > 
> > 
> It seems that the difficult point is when you go from no provider 
> available
> to one provider available. If that case is fixed then: No

That case also works.


Reinis


^ permalink raw reply	[flat|nested] 5+ messages in thread

end of thread, other threads:[~2014-01-22 10:02 UTC | newest]

Thread overview: 5+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2014-01-22  0:52 [gentoo-science] blas selection in lapack-reference François Bissey
2014-01-22  1:47 ` Steven Trogdon
2014-01-22  9:47   ` Reinis Danne
2014-01-22  9:54     ` François Bissey
2014-01-22 10:02       ` Reinis Danne

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox