On Fri, Oct 17, 2025 at 10:47:47AM +0200, Lorenzo Pieralisi wrote:
This is series is a follow up to [1] - with additional patches that are
addressing Rob's feedback (pcie-layerscape-gen4 was removed from the
kernel, Yay !) and other bits and bobs I noticed while staring at the code.
Patch (1) is a fix and technically we would like to get it in v6.18 please.
Patch (4) is compile-tested only, I can not run it on HW, I do not have it,
Scott, Ray please test it if you can.
v2 -> v3:
- Added additional patch to export of_msi_xlate()
- Merged Marc's ITS parent diff
- Added trailers, fixed commit logs
v2: https://lore.kernel.org/lkml/20251014095845.1310624-1-lpieralisi@kernel.org/ (local)
v1: https://lore.kernel.org/lkml/20250916091858.257868-1-lpieralisi@kernel.org/ (local)
[1] https://lore.kernel.org/lkml/20250916091858.257868-1-lpieralisi@kernel.org/ (local)
Cc: Sascha Bischoff <redacted>
Cc: Scott Branden <sbranden@broadcom.com>
Cc: Thomas Gleixner <redacted>
Cc: Bjorn Helgaas <bhelgaas@google.com>
Cc: Rob Herring <robh@kernel.org>
Cc: Ray Jui <rjui@broadcom.com>
Cc: Frank Li <Frank.Li@nxp.com>
Cc: Manivannan Sadhasivam <mani@kernel.org>
Cc: "Krzysztof Wilczyński" <kwilczynski@kernel.org>
Cc: Marc Zyngier <maz@kernel.org>
I tested epf msi doorbell, no regression. My test cover only using msi-map
code path.
For whole series
Reviewed-by: Frank Li <Frank.Li@nxp.com>
Lorenzo Pieralisi (5):
of/irq: Add msi-parent check to of_msi_xlate()
of/irq: Fix OF node refcount in of_msi_get_domain()
of/irq: Export of_msi_xlate() for module usage
PCI: iproc: Implement MSI controller node detection with
of_msi_xlate()
irqchip/gic-its: Rework platform MSI deviceID detection
drivers/irqchip/irq-gic-its-msi-parent.c | 91 ++++++------------------
drivers/of/irq.c | 43 +++++++++--
drivers/pci/controller/pcie-iproc.c | 22 ++----
3 files changed, 67 insertions(+), 89 deletions(-)
--
2.50.1