[PATCH 0/2] cxl/port: Robustness fixes for decoder enumeration

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Further testing of the decoder toplogy enumeration patches found cases
where the driver is too strict about what it accepts. First, there is no
expectation that the decoder's target list is valid when the decoder is
disabled. Make decoder_populate_targets() failures non-fatal on disabled
decoders. Second, if the decoder emits out-of-bounds / reserved values
at init warn and continue if at least one valid decoder was found. This
future-proofs the driver against changes to the interleave_ways
encoding, at least for continuing to operate decoders that conform to
current expectations.

Applies on top of:

https://lore.kernel.org/r/164298411792.3018233.7493009997525360044.stgit@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

---

Dan Williams (2):
      cxl/core/port: Fix / relax decoder target enumeration
      cxl/core/port: Handle invalid decoders


 drivers/cxl/acpi.c      |    2 +-
 drivers/cxl/core/hdm.c  |   36 ++++++++++++++++++++++++++++++------
 drivers/cxl/core/port.c |    5 ++++-
 3 files changed, 35 insertions(+), 8 deletions(-)



[Index of Archives]     [DMA Engine]     [Linux Coverity]     [Linux USB]     [Video for Linux]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]     [Greybus]

  Powered by Linux