All of lore.kernel.org
 help / color / mirror / Atom feed
* PRP1 not min page size aligned in read/write cmds
@ 2017-04-05  1:28 Nisha Miller
  2017-04-05 13:12 ` Keith Busch
  0 siblings, 1 reply; 2+ messages in thread
From: Nisha Miller @ 2017-04-05  1:28 UTC (permalink / raw)


Hi,

We have specified Min page size as 4K in our FW. So we expect that all
read/write commands sent to FW will operate on multiples of 4K blocks.
However we observe that sometimes we receive read/write commands where
the PRP1 address is not 4K aligned, e.g 0x84CD5CD50. Is this normal or
expected?

TIA
Nisha

^ permalink raw reply	[flat|nested] 2+ messages in thread

* PRP1 not min page size aligned in read/write cmds
  2017-04-05  1:28 PRP1 not min page size aligned in read/write cmds Nisha Miller
@ 2017-04-05 13:12 ` Keith Busch
  0 siblings, 0 replies; 2+ messages in thread
From: Keith Busch @ 2017-04-05 13:12 UTC (permalink / raw)


On Tue, Apr 04, 2017@06:28:30PM -0700, Nisha Miller wrote:
> We have specified Min page size as 4K in our FW. So we expect that all
> read/write commands sent to FW will operate on multiples of 4K blocks.
> However we observe that sometimes we receive read/write commands where
> the PRP1 address is not 4K aligned, e.g 0x84CD5CD50. Is this normal or
> expected?

Yes, the first PRP entry and any list pointer may have an offset. The
remaining entries can not. See section 4.3, figure 14 in NVMe 1.2.1 spec.

^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2017-04-05 13:12 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2017-04-05  1:28 PRP1 not min page size aligned in read/write cmds Nisha Miller
2017-04-05 13:12 ` Keith Busch

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.