Skip to content

Update kernel's PCI subsystem to v5.19

Myron Stowe requested to merge mstowe1/centos-stream-9:redhat/pci_v5.19rebase into main 
This series updates RHEL9's PCI subsystem with content from upstream v5.19 -
  Merge tag 'pci-v5.19-fixes-2' of git://git.kernel.org/../helgaas/pci
  https://lkml.org/lkml/2022/6/17/915
  commit 32efdbffff280b304bb2cd48844ee4166078e69c
  Merge: 93d17c1c8c1c a2b36ffbf5b6

  Merge tag 'pci-v5.19-fixes-1' of git://git.kernel.org/../helgaas/pci
  https://lkml.org/lkml/2022/6/2/753
  commit c399c85d60a4295bd911d366399a97df2865fb86
  Merge: 17d8e3d90b69 833e53a4ffe9


  Merge tag 'pci-v5.19-changes' of git://git.kernel.org/../helgaas/pci
  https://lkml.org/lkml/2022/5/26/820
  commit 3cc30140dbe2df9b5ac000898e0ae3d1df980f2c
  Merge: 8291eaafed36 32f479d05a44
  43 files changed, 1656 insertions(+), 758 deletions(-)

-v2  Alex Williamson caught that the revert of upstream commit 4c5e242d3e93
     was not a full revert: see a2b36ffbf5b6 x86/PCI: Revert "x86/PCI: 
     Clip only host bridge windows for E820 regions.  To fix such, both
     related upstream commits were added to this series. 

All patches within the series back-ported cleanly.


Bugzilla: https://bugzilla.redhat.com/show_bug.cgi?id=2118429
Depends: N/A
Tested: Nothing specifically targeted, just manual testing on a handful of
  machines looking for regressions of which none were seen; in general,
  if a system boots and runs with all known hardware active and usable,
  the PCI subsystem is behaving.

Signed-off-by: Myron Stowe <mstowe@redhat.com>
Edited by Myron Stowe

Merge request reports