On 25/02/10 at 11:29 +0100, Josip Rodin wrote: > On Mon, Feb 22, 2010 at 05:30:29PM +0100, Marc Brockschmidt wrote: > > > Any chance of upgrading one of the buildds to 2.6.32 to see if it > > > helps? > > > > DSA told me that they do not want to run any non-standard (aka, > > non-Debian stable) kernel on these machines, not even for short > > tests. I'm not sure that's the best way to handle this, but I can't > > change it. > > lebrun and schroeder had built many packages in the past with non-standard > kernels, indeed testing on them helped improve the standard upstream kernel, > and obviously we eventually do have to upgrade the buildds so it's a good > policy to find out if that's possible early enough. The one major data point > that you may be missing here is that a few months ago I got stuck trying to > to upgrade from the 2.6.28 to the 2.6.29 kernel built by new gcc on > schroeder. The same happened with .30 and .31; didn't test .32 yet but > there's no real indication that it would work. We've asked Dave Miller for > help, and he's promised to look into it. These heisenbuggy Ruby failures > sound like another thing you'd want him to look into.
Something else that puzzles me is that ruby1.9.1 1.9.1.376-1 (the version currently in testing) built fine on sparc back in August, when it was uploaded. See https://buildd.debian.org/fetch.cgi?pkg=ruby1.9.1;ver=1.9.1.376-1;arch=sparc;stamp=126038086 Marc, could you try to build 1.9.1.376-1 on lebrun and/or spontini? This could allow to pin down the problem to a toolchain regression (or change, at least). Thanks -- | Lucas Nussbaum | [email protected] http://www.lucas-nussbaum.net/ | | jabber: [email protected] GPG: 1024D/023B3F4F | -- To UNSUBSCRIBE, email to [email protected] with a subject of "unsubscribe". Trouble? Contact [email protected]
