libata-dev.git clean up and branch status

libata-dev.git clean up and branch status

[Jeff Garzik]

libata-dev.git was left a bit grotty when I went on vacation.  That has 
now been fixed.  The current branches are:

git://git.kernel.org/pub/scm/linux/kernel/git/jgarzik/libata-dev.git

master		Always vanilla Linus tree

ALL		Always the "exported for testing" meta-branch
		(currently == upstream + sil680-mmio + mv-ahci)

upstream	queued for 2.6.23
		Parent: master

mv-ahci		Marvell 6121/6141 AHCI clone SATA support
		Parent: upstream
		Status: Works, once NCQ & PCI MSI are disabled.
			Turn these off, then can go up.

mv-ahci-pata	Marvell 6121/6141 AHCI clone PATA support
		Parent: mv-ahci
		Status: Need to update for proper use of PATA
			controller commands (provides direct access to
			PATA device shadow registers)

mv-eh		sata_mv (50xx/60xx) conversion to new EH
		Parent: upstream
		Status: Required for hotplug, NCQ, etc.  Working, but
			currently NCQ disks fail (even if NCQ not used)

new-eh		pdc_adma, sata_qstor, sata_sx4 conversion to new EH
		Parent: upstream
		Status: All conversions complete...
			and completely untested.

sii-lbt		sata_sil Large [DMA] Block Transfer support
		Parent: upstream
		Status: Useful, but reported to fail in some cases

sil680-mmio	pata_sil680 MMIO support
		Parent: upstream
		Status: dissolved into some argument about MMIO flushing

Re: libata-dev.git clean up and branch status

[Alan Cox]

> sil680-mmio	pata_sil680 MMIO support
> 		Parent: upstream
> 		Status: dissolved into some argument about MMIO flushing

I thought we had figured out what needed sorting ?

Re: libata-dev.git clean up and branch status

[Alan Cox]

> mv-ahci-pata	Marvell 6121/6141 AHCI clone PATA support
> 		Parent: mv-ahci
> 		Status: Need to update for proper use of PATA
> 			controller commands (provides direct access to
> 			PATA device shadow registers)

What is the status on this - we've now had the regression where you
can't use all of a marvell chip for over 6 months ?

Alan