* CSB5 IDE does not fully support native mode??
@ 2005-05-24 20:26 evt
2005-05-26 14:24 ` Alan Cox
0 siblings, 1 reply; 4+ messages in thread
From: evt @ 2005-05-24 20:26 UTC (permalink / raw)
To: linux-kernel; +Cc: linux-ide
Can someone explain what the issue is and what I might need to
do to use csb5 ide in native mode? thanks.
- Eric van Tassell
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: CSB5 IDE does not fully support native mode??
2005-05-24 20:26 CSB5 IDE does not fully support native mode?? evt
@ 2005-05-26 14:24 ` Alan Cox
2005-05-26 16:55 ` Bjorn Helgaas
0 siblings, 1 reply; 4+ messages in thread
From: Alan Cox @ 2005-05-26 14:24 UTC (permalink / raw)
To: evt; +Cc: Linux Kernel Mailing List, linux-ide
On Maw, 2005-05-24 at 21:26, evt@texelsoft.com wrote:
> Can someone explain what the issue is and what I might need to
> do to use csb5 ide in native mode? thanks.
In order to keep the legacy world from falling to bits IDE and VGA have
some ugly hacks in the PCI spec.
In legacy mode an IDE device appears at the "old" standard IDE addresses
and uses an external IRQ pin wired to the ISA IRQ lines (14 or 15). In
native mode it behaves like a PCI device, honouring the PCI bars and
using the PCI INT lines.
>From a performance perspective the difference is essentially irrelevant
and from a wiring perspective its often decided irrevocably by the board
maker.
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: CSB5 IDE does not fully support native mode??
2005-05-26 14:24 ` Alan Cox
@ 2005-05-26 16:55 ` Bjorn Helgaas
2005-05-26 17:22 ` Alan Cox
0 siblings, 1 reply; 4+ messages in thread
From: Bjorn Helgaas @ 2005-05-26 16:55 UTC (permalink / raw)
To: Alan Cox; +Cc: evt, Linux Kernel Mailing List, linux-ide
On Thursday 26 May 2005 8:24 am, Alan Cox wrote:
> In order to keep the legacy world from falling to bits IDE and VGA have
> some ugly hacks in the PCI spec.
>
> In legacy mode an IDE device appears at the "old" standard IDE addresses
> and uses an external IRQ pin wired to the ISA IRQ lines (14 or 15). In
> native mode it behaves like a PCI device, honouring the PCI bars and
> using the PCI INT lines.
This has been niggling in my mind for a while -- in legacy mode,
the device should use IRQ 14/15. But I think we still call
pci_enable_device(), which sets up IRQ routing according to
the usual PCI rules. Should we be using pci_enable_device()
at all in legacy mode?
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: CSB5 IDE does not fully support native mode??
2005-05-26 16:55 ` Bjorn Helgaas
@ 2005-05-26 17:22 ` Alan Cox
0 siblings, 0 replies; 4+ messages in thread
From: Alan Cox @ 2005-05-26 17:22 UTC (permalink / raw)
To: Bjorn Helgaas; +Cc: evt, Linux Kernel Mailing List, linux-ide
On Iau, 2005-05-26 at 17:55, Bjorn Helgaas wrote:
> This has been niggling in my mind for a while -- in legacy mode,
> the device should use IRQ 14/15. But I think we still call
> pci_enable_device(), which sets up IRQ routing according to
> the usual PCI rules. Should we be using pci_enable_device()
> at all in legacy mode?
For all the other enables yes. The IRQ is an interesting case and I'd
have to look into the newer code to even guess - Bartlomiej has probably
has a much better idea as maintainer.
^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2005-05-26 17:24 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2005-05-24 20:26 CSB5 IDE does not fully support native mode?? evt
2005-05-26 14:24 ` Alan Cox
2005-05-26 16:55 ` Bjorn Helgaas
2005-05-26 17:22 ` Alan Cox
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).