ops->qc_defer not invoked on ata_exec_internal_sg() paths ?
The optional .qc_defer() methods don't seem to be called on the ata_exec_internal_sg() path.
At present, this is probably okay. But in the future, as we add functionality for link power management and hotplug polling, this could be a problem.
I think. Or is it possibly also a problem today for sata_send_pmp() and friends ?
Tejun?
Hello, Mark.
Quoted 9 lines · Mark Lord
The optional .qc_defer() methods don't seem to be called on the
ata_exec_internal_sg()path.At present, this is probably okay. But in the future, as we add functionality for link power management and hotplug polling, this could be a problem.
I think. Or is it possibly also a problem today for
sata_send_pmp()and friends ?
That's intentional and okay. Currently, EH can only issue one non-NCQ command at a time and dynamic link PM and hotplug polling shouldn't be done via EH anyway. EH is way too big a hammer for those.
Thanks.
Signature · 1 line
tejun
Quoted 6 lines · On Mon, 25 Feb 2008 19:20:59 -0500 Mark Lord <[email protected]> wrote:
The optional .qc_defer() methods don't seem to be called on the
ata_exec_internal_sg()path.At present, this is probably okay. But in the future, as we add functionality for link power management and hotplug polling, this could be a problem.
It becomes a problem for me certainly if we ever issue anything via exec_internal_sg while another command may be outstanding anywhere on the same *host*. Right now its ok.
Alan
Quoted 6 lines · Alan Cox
Quoted 3 lines · On Mon, 25 Feb 2008 19:20:59 -0500 Mark Lord <[email protected]> wrote:
The optional .qc_defer() methods don't seem to be called on the
ata_exec_internal_sg()path.
Quoted 7 lines
Quoted 3 lines
At present, this is probably okay. But in the future, as we add functionality for link power management and hotplug polling, this could be a problem.
It becomes a problem for me certainly if we ever issue anything via
exec_internal_sgwhile another command may be outstanding anywhere on the same *host*. Right now its ok.
..
Yeah, for serializing a pair of ATA channels on the same host chip.
-ml