On Thu, 30 Dec 2010, Arnd Bergmann wrote:

quoted text
> > Doesn't this patch series break the 'bisectable' rule for patches?
> > I think you can get by with adding a CONFIG_EXPERT selects CONFIG_EMBEDDED
> > in the first patch and the last patch removing that same selects.
> 
> How about leaving it that way during the merge window even? That would
> help people sending git pull requests that add new dependencies on
> CONFIG_EMBEDDED. David can submit another patch that removes the
> temporary CONFIG_EMBEDDED option along with any new users in -rc2
> or for 2.6.39 then.
> 

Ok, so there are a few pending EMBEDDED dependency additions in linux-next 
as of next-20101230, which should be a pretty good indicator of what type 
of conflict we might risk during the merge window.

 - three new defconfig files:

	89ba4d12: arch/arm/configs/ag5evm_defconfig
	6cfce27c: arch/arm/configs/omap1_defconfig 
	d58f0967: arch/blackfin/configs/BF561-EZKIT-SMP_defconfig 

 - 3ce05168 (drm/kms: load fbcon from drm_kms_helper)

 - b595076a (tree-wide: fix comment/printk typos)

Instead of CONFIG_EXPERT select CONFIG_EMBEDDED, we could add a ninth 
patch which does CONFIG_EMBEDDED select CONFIG_EXPERT and tag it in some 
way so we know it's only temporary for the merge window to find new 
CONFIG_EMBEDDED entries.

Andrew, this would be going through your -mm tree anyway, so how would you 
prefer to merge it?
--
unsubscribe notice
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/