ACK/Cmnt: [SRU][P/N/J:linux-azure][PULL] Additional MANA patch support

Philip Cox philip.cox at canonical.com
Fri Jul 25 18:56:39 UTC 2025


On 2025-07-14 3:36 p.m., John Cabaj wrote:
> BugLink: https://bugs.launchpad.net/bugs/2115309
> BugLink: https://bugs.launchpad.net/ubuntu/bugs/2115356

^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^


This buglink doesn't work.  You probably meant 
https://bugs.launchpad.net/bugs/2115356



>
> SRU Justification
>
> [Impact]
>
> * Microsoft requested MANA patches for the following:
>     * Better spread IRQs per cpu
>     * Allocate MSI-X vectors in PCI dynamically
>     * Represent packets transmitted after GSO processing
>     * Support holes in device list reply msg
>     * Support multi vports on bare-metal
>     * Action on receiving the special EQE notification
>
> [Fix]
>
> * Plucky:
>     * 1 clean cherry-pick from upstream
>     * 290e5d3c49f6: "net: mana: Add support for Multi Vports on Bare 
> metal"
>     * 5 clean cherry-picks from linux-next
>     * 5da8a8b8090b: "PCI/MSI: Export pci_msix_prepare_desc() for 
> dynamic MSI-X allocations"
>     * ad518f2557b9: "PCI: hv: Allow dynamic MSI-X vector allocation"
>     * 4607617af1b4: "net: mana: explain irq_setup() algorithm"
>     * 845c62c543d6: "net: mana: Allow irq_setup() to skip cpus for 
> affinity"
>     * 7399ef984022: "net: mana: Set tx_packets to post gso processing 
> packet count"
>     * 2 backports from linux-next
>     * 755391121038: "net: mana: Allocate MSI-X vectors dynamically"
>     * 7768c5f41733: "net: mana: Add handler for hardware servicing 
> events"
>
> * Noble:
>     * 6 clean cherry-picks from upstream
>     * c1f5204efcbc: "cpumask: add cpumask_weight_andnot()"
>     * dcee228078c3: "cpumask: define cleanup function for cpumasks"
>     * 91bfe210e196: "net: mana: add a function to spread IRQs per CPUs"
>     * 8afefc361209: "net: mana: Assigning IRQ affinity on HT cores"
>     * 2fc8a346625e: "net: mana: Support holes in device list reply msg"
>     * 290e5d3c49f6: "net: mana: Add support for Multi Vports on Bare 
> metal"
>     * 5 clean cherry-picks from linux-next
>     * 5da8a8b8090b: "PCI/MSI: Export pci_msix_prepare_desc() for 
> dynamic MSI-X allocations"
>     * ad518f2557b9: "PCI: hv: Allow dynamic MSI-X vector allocation"
>     * 845c62c543d6: "net: mana: explain irq_setup() algorithm"
>     * 845c62c543d6: "net: mana: Allow irq_setup() to skip cpus for 
> affinity"
>     * 7399ef984022: "net: mana: Set tx_packets to post gso processing 
> packet count"
>     * 2 backports from linux-next
>     * 755391121038: "net: mana: Allocate MSI-X vectors dynamically"
>     * 7768c5f41733: "net: mana: Add handler for hardware servicing 
> events"
>
> * Jammy:
>     * 2 clean cherry-picks from upstream
>     * 2fc8a346625e: "net: mana: Support holes in device list reply msg"
>     * 290e5d3c49f6: "net: mana: Add support for Multi Vports on Bare 
> metal"
>     * 2 clean cherry-picks from linux-next
>     * 7399ef984022: "net: mana: Set tx_packets to post gso processing 
> packet count"
>     * 7768c5f41733: "net: mana: Add handler for hardware servicing 
> events"


This last jammy commit has the same broken buglink that this cover 
letter has.  Fix the buglink, and then this is all

Acked-by: Philip Cox <philip.cox at canonical.com>



>
> [Test Plan]
>
> * Compile tested
> * Microsoft tested
>
> [Regression potential]
>
> * Changes are largely localized to MANA and Hyper-V, and either 
> already upstream or queued for upstream in the next merge window. 
> Possible regressions could result in unexpected behavior of MANA driver.
>
> [Other info]
>
> * SF#00413058
> * SF#00413050
> * SF#00413245
> * SF#00413047
> * SF#00411917
> * SF#00413476
>



More information about the kernel-team mailing list