* [irq-msi] questions about irq msi_desc usage
@ 2025-12-17 1:58 Chao Xie
2025-12-17 20:41 ` Thomas Gleixner
0 siblings, 1 reply; 2+ messages in thread
From: Chao Xie @ 2025-12-17 1:58 UTC (permalink / raw)
To: linux-kernel; +Cc: tglx
I intend to upstream a driver for a new bus, which utilizes MSI IRQ.
I have some questions regarding how to implement MSI IRQ support for it:
1. To define my own private data structure, should I follow the PCI MSI
approach and add a structure like struct my_private_intr_desc to
the union?
struct msi_desc {
...
union {
struct pci_msi_desc pci;
struct msi_desc_data data;
+++ struct private_intr_desc private_intr;
};
};
I think I can make use of data.docckie.ptr to point to my private data
structure, but I do not find any example in kernel, and not sure whether
It is right or not.
2. If I use data.dcookie.ptr to point to my private data structure,
should this association be initialized in msi_domain_ops.set_desc()?
3. The Type 1 interrupt of this bus is similar to MSI.
Specifically, its configuration space includes registers such as
interrupt_enable, interrupt_mask, and interrupt_data, with the following
specifications:
• Interrupt Enable: The software configures this register to
control whether the interrupt function of the Entity is enabled.
• Interrupt Data: The software configures the start number of
interrupt vectors into this register. When the Entity reports an
interrupt, it shall follow the rules below:
(1) When one interrupt vector is enabled, the interrupt data
reported by the Entity shall be X, where X is the value
configured in the Interrupt Data register.
(2) When N interrupt vectors are enabled, the interrupt data
reported by the Entity shall range from X to X+N-1, and
the value shall correspond to the sequence of interrupt
vectors. X is the value configured in the Interrupt Data
register.
• Interrupt Address: The software configures the interrupt
address allocated by the interrupt subsystem into this register.
• Interrupt Mask: The software configures this register to
control whether the interrupt reporting corresponding to the
Entity is masked. Its 32 bits correspond to 32 interrupt vector
numbers respectively.
In this case, when allocating msi_desc, should I follow the PCI MSI
model where one msi_desc represents multiple interrupts
(i.e.,set msi_desc.nvec_used = n)? If this is the correct approach,
is itfeasible to add a judgment similar to PCI MSI in
msi_domain_get_virq() as follows:
if (dev_is_pci(dev) && domid == MSI_DEFAULT_DOMAIN)
devmsi = to_pci_dev(dev)->msi_enabled;
+++ if (dev_is_my_private(dev) && domid == MSI_DEFAULT_DOMAIN)
+++ devmsi = to_my_private_entity(dev)->is_type1;
^ permalink raw reply [flat|nested] 2+ messages in thread
* Re: [irq-msi] questions about irq msi_desc usage
2025-12-17 1:58 [irq-msi] questions about irq msi_desc usage Chao Xie
@ 2025-12-17 20:41 ` Thomas Gleixner
0 siblings, 0 replies; 2+ messages in thread
From: Thomas Gleixner @ 2025-12-17 20:41 UTC (permalink / raw)
To: Chao Xie, linux-kernel
On Wed, Dec 17 2025 at 09:58, Chao Xie wrote:
> 1. To define my own private data structure, should I follow the PCI MSI
> approach and add a structure like struct my_private_intr_desc to
> the union?
No.
> I think I can make use of data.docckie.ptr to point to my private data
> structure, but I do not find any example in kernel, and not sure whether
> It is right or not.
Yes it is the right thing to do if you really need private data for your
driver. You did not explain what this data is used for and why it needs
to be accessible through msi_desc, so I can't tell.
> 2. If I use data.dcookie.ptr to point to my private data structure,
> should this association be initialized in msi_domain_ops.set_desc()?
prepare_desc() is probably the right place especially if you need access
to the domain itself.
> 3. The Type 1 interrupt of this bus is similar to MSI.
Why are hardware people insisting on repeating the same mistakes over
and over?
> In this case, when allocating msi_desc, should I follow the PCI MSI
> model where one msi_desc represents multiple interrupts
> (i.e.,set msi_desc.nvec_used = n)?
That's the only way you can guarantee consecutive Linux interrupt
numbers for the block you allocate.
> If this is the correct approach,
> is itfeasible to add a judgment similar to PCI MSI in
> msi_domain_get_virq() as follows:
> if (dev_is_pci(dev) && domid == MSI_DEFAULT_DOMAIN)
> devmsi = to_pci_dev(dev)->msi_enabled;
> +++ if (dev_is_my_private(dev) && domid == MSI_DEFAULT_DOMAIN)
> +++ devmsi = to_my_private_entity(dev)->is_type1;
No. This wants to be generalized.
Thanks,
tglx
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2025-12-17 20:41 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2025-12-17 1:58 [irq-msi] questions about irq msi_desc usage Chao Xie
2025-12-17 20:41 ` Thomas Gleixner
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.