воскресенье

[Bug 2146720] [NEW] Request for NVMe Hot-Insert Support – PCIe System Firmware Intermediary (SFI) Handling in Ubuntu 26.04

Private bug reported:

NVMe hot-insert enables dynamic addition of NVMe storage devices into a
running system without requiring a reboot. In advanced server platforms,
especially those with complex PCIe topologies (switches, backplanes,
multi-host configurations), hotplug operations can be coordinated
through a System Firmware Intermediary (SFI) model.

In the SFI-based flow, system firmware acts as an intermediary between
hardware events and the operating system. When an NVMe device is
inserted, firmware detects the event, performs initial validation and
platform-specific policy enforcement (e.g., power control, slot
management, link training), and then notifies the OS with a structured
event.

This approach ensures consistent handling of hotplug events across
platforms, especially where direct OS-first handling may not be
sufficient due to hardware dependencies or platform constraints. It also
enables integration with platform management components such as BMCs and
backplane controllers.

In the Linux kernel, NVMe hotplug is supported through PCIe hotplug
(pciehp) and NVMe subsystems. However, SFI-mediated hot-insert flows
require enhanced coordination between firmware notifications (e.g.,
ACPI, platform-specific interfaces) and OS device enumeration, driver
binding, and error handling. Improved support is needed to ensure
reliable and seamless NVMe hot-insert operations in firmware-
intermediated environments.

Feature Request:
Requested details to be enabled on OS:
  Enable support for SFI-based NVMe hot-insert event handling. 
  Integrate firmware-generated hotplug notifications (ACPI, mailbox, or vendor-specific) with OS hotplug framework. 
  Ensure proper PCIe enumeration and NVMe driver binding upon device insertion. 
  Support coordinated power-on, link training, and device initialization flows via firmware. 
  Enhance pciehp and NVMe subsystems to handle firmware-mediated events. 
  Provide visibility into hotplug events and states via sysfs/debugfs. 
  Support error handling and recovery during hot-insert operations. 
  Enable integration with BMC/management frameworks for coordinated control. 
  Support hot-insert across PCIe switches and complex topologies. 
  Provide validation and debugging tools for SFI hotplug flows. 
  Ensure compatibility with PCIe Gen5/Gen6 and future NVMe devices. 
  Document supported workflows, configuration, and platform dependencies.

Business Justification:
 Enables seamless NVMe device insertion without system downtime.
 Improves reliability and consistency of hotplug operations in complex platforms.
 Enhances coordination between firmware and OS for device lifecycle management.
 Supports scalable storage expansion in enterprise and cloud environments.
 Reduces operational complexity in managing storage infrastructure.
 Aligns with modern data center requirements for dynamic and composable storage.

References:
  PCI-SIG PCIe Specification (Hotplug, AER, DPC) 
  NVMe Specification (Hotplug and Device Initialization) 
  ACPI Specification (Hotplug Event Notification) 
  Linux Kernel PCIe Hotplug and NVMe Subsystem Documentation

** Affects: linux (Ubuntu)
Importance: Undecided
Status: New

** Information type changed from Public to Private

--
You received this bug notification because you are subscribed to linux
in Ubuntu.
Matching subscriptions: Bgg, Bmail, Nb
https://bugs.launchpad.net/bugs/2146720

Title:
Request for NVMe Hot-Insert Support – PCIe System Firmware
Intermediary (SFI) Handling in Ubuntu 26.04

Status in linux package in Ubuntu:
New

Bug description:
NVMe hot-insert enables dynamic addition of NVMe storage devices into
a running system without requiring a reboot. In advanced server
platforms, especially those with complex PCIe topologies (switches,
backplanes, multi-host configurations), hotplug operations can be
coordinated through a System Firmware Intermediary (SFI) model.

In the SFI-based flow, system firmware acts as an intermediary between
hardware events and the operating system. When an NVMe device is
inserted, firmware detects the event, performs initial validation and
platform-specific policy enforcement (e.g., power control, slot
management, link training), and then notifies the OS with a structured
event.

This approach ensures consistent handling of hotplug events across
platforms, especially where direct OS-first handling may not be
sufficient due to hardware dependencies or platform constraints. It
also enables integration with platform management components such as
BMCs and backplane controllers.

In the Linux kernel, NVMe hotplug is supported through PCIe hotplug
(pciehp) and NVMe subsystems. However, SFI-mediated hot-insert flows
require enhanced coordination between firmware notifications (e.g.,
ACPI, platform-specific interfaces) and OS device enumeration, driver
binding, and error handling. Improved support is needed to ensure
reliable and seamless NVMe hot-insert operations in firmware-
intermediated environments.

Feature Request:
Requested details to be enabled on OS:
  Enable support for SFI-based NVMe hot-insert event handling. 
  Integrate firmware-generated hotplug notifications (ACPI, mailbox, or vendor-specific) with OS hotplug framework. 
  Ensure proper PCIe enumeration and NVMe driver binding upon device insertion. 
  Support coordinated power-on, link training, and device initialization flows via firmware. 
  Enhance pciehp and NVMe subsystems to handle firmware-mediated events. 
  Provide visibility into hotplug events and states via sysfs/debugfs. 
  Support error handling and recovery during hot-insert operations. 
  Enable integration with BMC/management frameworks for coordinated control. 
  Support hot-insert across PCIe switches and complex topologies. 
  Provide validation and debugging tools for SFI hotplug flows. 
  Ensure compatibility with PCIe Gen5/Gen6 and future NVMe devices. 
  Document supported workflows, configuration, and platform dependencies.

Business Justification:
 Enables seamless NVMe device insertion without system downtime.
 Improves reliability and consistency of hotplug operations in complex platforms.
 Enhances coordination between firmware and OS for device lifecycle management.
 Supports scalable storage expansion in enterprise and cloud environments.
 Reduces operational complexity in managing storage infrastructure.
 Aligns with modern data center requirements for dynamic and composable storage.

References:
  PCI-SIG PCIe Specification (Hotplug, AER, DPC) 
  NVMe Specification (Hotplug and Device Initialization) 
  ACPI Specification (Hotplug Event Notification) 
  Linux Kernel PCIe Hotplug and NVMe Subsystem Documentation

To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/2146720/+subscriptions

Комментариев нет:

Отправить комментарий