Re: [PATCH 1/4] libata: implement ata_exec_internal()

[Jeff Garzik <jgarzik@pobox.com>]

Tejun Heo wrote:
> Oh... Sorry, all patches are against the ALL branch.  Do you want me to 
> regenerate them against upstream?

Yes, please.


> Also, I'm currently working on a series of EH improvement patches 
> against the ALL branch.  Would generating patches against upstream be 
> better?

Unfortunately, generating patches against the ALL branch makes a lot of 
sense -- but it's also the worst choice you could make.

Generating patches against the ALL branch means that your patch becomes 
dependent on -all- branches in libata-dev.git, even the branches that 
aren't going upstream for quite a while...  which is generally not what 
you want at all.

ALL is largely for testing and automated push to -mm, but not really for 
developers.

Sometimes you will run into cases where I slacked off on the merge of 
upstream+irq-pio+etc and created a bug, but most of the time the merge 
is easy.  If there are ALL-specific patches, you are welcome to submit 
them separately, with the recogition that any such ALL-specific patch 
would likely be dropped and recreated each time ALL gets blown away, and 
recreated.

	Jeff