linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* Why MSI is limited to 32 interrupts maximum
@ 2016-10-15 12:37 Bharat Kumar Gogada
  2016-10-15 13:14 ` Marc Zyngier
  0 siblings, 1 reply; 4+ messages in thread
From: Bharat Kumar Gogada @ 2016-10-15 12:37 UTC (permalink / raw)
  To: linux-pci, linux-kernel; +Cc: Bjorn Helgaas, arnd, Marc Zyngier

Hi

Can anyone tell why MSI interrupts are limited to maximum 32 interrupts, even though we have 16bit message data register ?

Regards,
Bharat


This email and any attachments are intended for the sole use of the named recipient(s) and contain(s) confidential information that may be proprietary, privileged or copyrighted under applicable law. If you are not the intended recipient, do not read, copy, or forward this email message or any attachments. Delete this email message and any attachments immediately.

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: Why MSI is limited to 32 interrupts maximum
  2016-10-15 12:37 Why MSI is limited to 32 interrupts maximum Bharat Kumar Gogada
@ 2016-10-15 13:14 ` Marc Zyngier
  2016-10-15 13:30   ` Bharat Kumar Gogada
  0 siblings, 1 reply; 4+ messages in thread
From: Marc Zyngier @ 2016-10-15 13:14 UTC (permalink / raw)
  To: Bharat Kumar Gogada; +Cc: linux-pci, linux-kernel, Bjorn Helgaas, arnd

On Sat, 15 Oct 2016 12:37:55 +0000
Bharat Kumar Gogada <bharat.kumar.gogada@xilinx.com> wrote:

Hi Bharat,

> Can anyone tell why MSI interrupts are limited to maximum 32
> interrupts, even though we have 16bit message data register ?

That's the very definition of the original PCI MSI: Up to 32
consecutive interrupts per function, and a single doorbell address.
MSI-X lifts that restriction and offers up to 2048 interrupts that do
not have to be contiguous can target individual doorbells. Assuming
your MSI controller advertises MSI-X support and that your devices do
support it as well, you'll be able to enjoy those.

> 
> Regards,
> Bharat
> 
> 
> This email and any attachments are intended for the sole use of the
> named recipient(s) and contain(s) confidential information that may
> be proprietary, privileged or copyrighted under applicable law. If
> you are not the intended recipient, do not read, copy, or forward
> this email message or any attachments. Delete this email message and
> any attachments immediately.
> 

Please fix your email not to carry such disclaimer (or use a private
email address if you can't work around your company policy). Posting
confidential information on a public mailing list feels a bit silly.

Thanks,

	M.
-- 
Without deviation from the norm, progress is not possible.

^ permalink raw reply	[flat|nested] 4+ messages in thread

* RE: Why MSI is limited to 32 interrupts maximum
  2016-10-15 13:14 ` Marc Zyngier
@ 2016-10-15 13:30   ` Bharat Kumar Gogada
  2016-10-15 16:39     ` okaya
  0 siblings, 1 reply; 4+ messages in thread
From: Bharat Kumar Gogada @ 2016-10-15 13:30 UTC (permalink / raw)
  To: Marc Zyngier; +Cc: linux-pci, linux-kernel, Bjorn Helgaas, arnd

 > On Sat, 15 Oct 2016 12:37:55 +0000
> Bharat Kumar Gogada <bharat.kumar.gogada@xilinx.com> wrote:
> 
> Hi Bharat,
> 
> > Can anyone tell why MSI interrupts are limited to maximum 32
> > interrupts, even though we have 16bit message data register ?
> 
> That's the very definition of the original PCI MSI: Up to 32 consecutive interrupts
> per function, and a single doorbell address.
> MSI-X lifts that restriction and offers up to 2048 interrupts that do not have to
> be contiguous can target individual doorbells. Assuming your MSI controller
> advertises MSI-X support and that your devices do support it as well, you'll be
> able to enjoy those.
> 

Thanks Marc, so then only 5 bits of message data register will be used, so what is the purpose rest of the bits ?

> >
> > Regards,
> > Bharat
> >
> >
> > This email and any attachments are intended for the sole use of the
> > named recipient(s) and contain(s) confidential information that may be
> > proprietary, privileged or copyrighted under applicable law. If you
> > are not the intended recipient, do not read, copy, or forward this
> > email message or any attachments. Delete this email message and any
> > attachments immediately.
> >
> 
> Please fix your email not to carry such disclaimer (or use a private email address
> if you can't work around your company policy). Posting confidential information
> on a public mailing list feels a bit silly.
> 
> Thanks,
> 
> 	M.
> --
> Without deviation from the norm, progress is not possible.

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: Why MSI is limited to 32 interrupts maximum
  2016-10-15 13:30   ` Bharat Kumar Gogada
@ 2016-10-15 16:39     ` okaya
  0 siblings, 0 replies; 4+ messages in thread
From: okaya @ 2016-10-15 16:39 UTC (permalink / raw)
  To: Bharat Kumar Gogada
  Cc: Marc Zyngier, linux-pci, linux-kernel, Bjorn Helgaas, arnd,
	linux-pci-owner

On 2016-10-15 06:30, Bharat Kumar Gogada wrote:
>> On Sat, 15 Oct 2016 12:37:55 +0000
>> Bharat Kumar Gogada <bharat.kumar.gogada@xilinx.com> wrote:
>> 
>> Hi Bharat,
>> 
>> > Can anyone tell why MSI interrupts are limited to maximum 32
>> > interrupts, even though we have 16bit message data register ?
>> 
>> That's the very definition of the original PCI MSI: Up to 32 
>> consecutive interrupts
>> per function, and a single doorbell address.
>> MSI-X lifts that restriction and offers up to 2048 interrupts that do 
>> not have to
>> be contiguous can target individual doorbells. Assuming your MSI 
>> controller
>> advertises MSI-X support and that your devices do support it as well, 
>> you'll be
>> able to enjoy those.
>> 
> 
> Thanks Marc, so then only 5 bits of message data register will be
> used, so what is the purpose rest of the bits ?
> 

Data field is used to indicate which interrupt number is used. It does 
not need to start at 0.


You have 32 consecutive interrupts starting from some arbitrary base.


>> >
>> > Regards,
>> > Bharat
>> >
>> >
>> > This email and any attachments are intended for the sole use of the
>> > named recipient(s) and contain(s) confidential information that may be
>> > proprietary, privileged or copyrighted under applicable law. If you
>> > are not the intended recipient, do not read, copy, or forward this
>> > email message or any attachments. Delete this email message and any
>> > attachments immediately.
>> >
>> 
>> Please fix your email not to carry such disclaimer (or use a private 
>> email address
>> if you can't work around your company policy). Posting confidential 
>> information
>> on a public mailing list feels a bit silly.
>> 
>> Thanks,
>> 
>> 	M.
>> --
>> Without deviation from the norm, progress is not possible.
> --
> To unsubscribe from this list: send the line "unsubscribe linux-pci" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

^ permalink raw reply	[flat|nested] 4+ messages in thread

end of thread, other threads:[~2016-10-15 16:39 UTC | newest]

Thread overview: 4+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2016-10-15 12:37 Why MSI is limited to 32 interrupts maximum Bharat Kumar Gogada
2016-10-15 13:14 ` Marc Zyngier
2016-10-15 13:30   ` Bharat Kumar Gogada
2016-10-15 16:39     ` okaya

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).