On Wed, 13 May 2015 19:43:09 +0200, Luc Maisonobe wrote:
Le 13/05/2015 11:35, Gilles a écrit :
Hello.
On Wed, 13 May 2015 11:12:16 +0200, Luc Maisonobe wrote:
Hi all,
As Jenkins failures appeared again, I restarted work on
FastMath.pow.
I'll keep you informed as I make progess.
Did someone see those failures anywhere else other than on the
identified machine used for CI?
I did not reproduce them myselves, and I guess most of us didn't
succeed
either (lets remember last year when Thomas and Sebb tried to fix the
exp function).
However, the configuration is not a rare one (Ubuntu, Java 7, Sun
JVM),
so I fear this bug, even if rare, can hit several people.
How much damage can it do, apart from a false positive on a unit test?
Isn't it rather the configuration that is faulty, rather than the CM
code?
Best,
Gilles
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
