On 06/07/2012 07:11 PM, Alexander Graf wrote:
On 08.06.2012, at 00:48, Jon Masters wrote:
On 06/07/2012 02:05 PM, Peter Robinson wrote:
On Thu, Jun 7, 2012 at 6:09 PM, Steve McIntyre
Fedora
Fedora 17 has been released for primary platforms, but not for ARM yet. Still some remaining work.
- OMAP4 "doesn't work" (crashes), likely due to reliance on pure
upstream OMAP support rather than TI branches. Other distros have used TI patches.
It now works fine. It's a bug in the CONFIG_OMAP4_ERRATA_I688 that arrived in the 3.3 kernel. Disabled that and it now works fine.
Right. It was interesting, that we turn out to be almost the only folks using an upstream kernel. I talked with some TI folks about how to get better upstream-only testing with configs like Fedora. Let's followup on that. This might be a config issue, but hopefully we can get some others to help test configs and find this kind of thing more quickly.
- Might be more similar issues around other SoCs that lack support in
upstream kernel.
The SoC's we support (omap, tegra, imx, kirkwood, highbank and vexpress) all work fine with a few minor patches to fix build issues and bugs. Most of the patches are just a few lines.
Indeed. But the point is still there. Lots of other folks are using non-upstream trees as their source, which is a difference, and so we wanted to convey that Fedora is taking a different (and I think ultimately the right one, but I'm biased) approach.
Not sure I agree here. OpenSUSE is 100% upstream
Good to know. I don't want to make this about distro X or Y (even on cross-disto!) but more about the notion. It's clear that there is some difference in opinion and usage and I think the main thing I want to convey is that this is a reality. So it's not a "blame" thing, it's just that there's a difference in approach and more than I thought was the case some of the sub-trees are getting more direct use out there.
Again, it all feeds back into getting stuff tested with non-defconfig.
Jon.