mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
* No sharing IRQ broken board problem
@ 2005-11-02 21:06 tcrix
  2005-11-02 21:13 ` Lennart Sorensen
  2005-11-02 21:32 ` linux-os (Dick Johnson)
  0 siblings, 2 replies; 3+ messages in thread
From: tcrix @ 2005-11-02 21:06 UTC (permalink / raw)
  To: Linux Kernel Mailing List

I have a pci board in development that works but..
it can not share the interrupt line.  

Has someone hacked through the problem of reserving one of the inta, intb, .. 's for a single device?  I would love to see how you did so I could continue on with my driver while I wait for the !@$#!@ hardware guys to fix the board. 

Thanks for the help,
Tom Rix





^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: No sharing IRQ broken board problem
  2005-11-02 21:06 No sharing IRQ broken board problem tcrix
@ 2005-11-02 21:13 ` Lennart Sorensen
  2005-11-02 21:32 ` linux-os (Dick Johnson)
  1 sibling, 0 replies; 3+ messages in thread
From: Lennart Sorensen @ 2005-11-02 21:13 UTC (permalink / raw)
  To: tcrix; +Cc: Linux Kernel Mailing List

On Wed, Nov 02, 2005 at 09:06:53PM +0000, tcrix@att.net wrote:
> I have a pci board in development that works but..
> it can not share the interrupt line.  
> 
> Has someone hacked through the problem of reserving one of the inta, intb, .. 's for a single device?  I would love to see how you did so I could continue on with my driver while I wait for the !@$#!@ hardware guys to fix the board. 

As far as I understand the PCI specs, your device is currently NOT a PCI
device.

Support for shared IRQs on PCI is as far as I can tell not optional.
Short of designing your own mainboard there is no way to prevent the
interrupt lines from being joined.  One one board I use there are 6 PCI
devices, with 3 on INT-A and 3 on INT-B (The SBC doesn't have INT-C and
INT-D).  On my desktop machine, INT A is the same on slot 1 and 5, 2,
and 6 and onyl 3 and 4 have INT-A to themselves.  Of course this doesn't
prevent the bios from assigning the same IRQ to INT-A on multiple sets
as well.  I have run a PPro some years ago with a single IRQ shared
among all the PCI devices in the system.  Neither linux nor windows had
a problem with that and neither did any of the many PCI devices.

I think you are stuck waiting for the hardware guys to give you a PCI
device that is really a PCI device.  If you are lucky you can find one
slot in your machine that doesn't share INT-A with any other PCI device,
although on most modern machines there are so many PCI devices that it
is imposible to find one that isn't shared.

Len Sorensen

^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: No sharing IRQ broken board problem
  2005-11-02 21:06 No sharing IRQ broken board problem tcrix
  2005-11-02 21:13 ` Lennart Sorensen
@ 2005-11-02 21:32 ` linux-os (Dick Johnson)
  1 sibling, 0 replies; 3+ messages in thread
From: linux-os (Dick Johnson) @ 2005-11-02 21:32 UTC (permalink / raw)
  To: tcrix; +Cc: Linux Kernel Mailing List


On Wed, 2 Nov 2005 tcrix@att.net wrote:

> I have a pci board in development that works but..
> it can not share the interrupt line.
>

It's on the PCI bus. It needs to be an "open collector" type.
Sharing or not is really not a software issue! The only
kind of hardware problem that could cause "shared interrupt"
line problems is the designer fails to provide the correct
kind of interface driver. In that case, just "dead-bug" a
FET (or two) on the board, near the connector, to convert
the driven bit to an open-collector (drain) bit.

> Has someone hacked through the problem of reserving one of the inta, intb,
> .. 's for a single device?  I would love to see how you did so I could
> continue on with my driver while I wait for the !@$#!@ hardware guys to fix
> the board.
>
> Thanks for the help,
> Tom Rix
> -

Cheers,
Dick Johnson
Penguin : Linux version 2.6.13.4 on an i686 machine (5589.55 BogoMips).
Warning : 98.36% of all statistics are fiction.
.

****************************************************************
The information transmitted in this message is confidential and may be privileged.  Any review, retransmission, dissemination, or other use of this information by persons or entities other than the intended recipient is prohibited.  If you are not the intended recipient, please notify Analogic Corporation immediately - by replying to this message or by sending an email to DeliveryErrors@analogic.com - and destroy all copies of this information, including any attachments, without reading or disclosing them.

Thank you.

^ permalink raw reply	[flat|nested] 3+ messages in thread

end of thread, other threads:[~2005-11-02 21:32 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2005-11-02 21:06 No sharing IRQ broken board problem tcrix
2005-11-02 21:13 ` Lennart Sorensen
2005-11-02 21:32 ` linux-os (Dick Johnson)

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox

all inboxes | Powered by JetHome®