[v4,0/6] hw/cxl: Poison get, inject, clear

Message

Jonathan Cameron March 3, 2023, 3:09 p.m. UTC

Note there are several series ahead of this one and in particular
the RAS error injection series needs some QAPI review.
The QAPI stuff in this patch is similar but in essence very similar
to what we have in that series.

Whilst I'm an always an optimist, this may well end up as 8.1 material
now.

Chance since v3: Thanks to Ira for review.
- Expanded the 'source' mask to allow for vendor defined source.
  Note this is just to simplify potential future support for injecting
  poison with that source. As of today there is no way of doing it.
- Dropped an overly paranoid overflow check in the clear poison handling.
- Ensure that we leave the poison list in a sane state in the overflow
  during clear case.  Previously it ended up one entry too big.
  Note that to test those overflow cases, I changed the limit to 4 entries
  to make them easier to trigger.
- Fix an off by one in the edge of the volatie region when clearning.
  Copy of a previously fixed bug found in the volatile memory support
  series that is a precursor of this one.

Based on following series (in order)
1. [PATCH v4 00/10] hw/cxl: CXL emulation cleanups and minor fixes for upstream
(currently in staging, so hopefully will land in upstream shortly!)
2. [PATCH v6 0/8] hw/cxl: RAS error emulation and injection
3. [PATCH v2 0/2] hw/cxl: Passthrough HDM decoder emulation
4. [PATCH v4 0/2] hw/mem: CXL Type-3 Volatile Memory Support

Based on: Message-Id: [e-mail]
Based-on: Message-id: [e-mail]
Based-on: Message-id: [e-mail]
Based-on: Message-id: [e-mail]

The series supports:
1) Injection of variable length poison regions via QMP (to fake real
   memory corruption and ensure we deal with odd overflow corner cases
   such as clearing the middle of a large region making the list overflow
   as we go from one long entry to two smaller entries.
2) Read of poison list via the CXL mailbox.
3) Injection via the poison injection mailbox command (limited to 64 byte
   entries - spec constraint)
4) Clearing of poison injected via either method.

The implementation is meant to be a valid combination of impdef choices
based on what the spec allowed. There are a number of places where it could
be made more sophisticated that we might consider in future:
* Fusing adjacent poison entries if the types match.
* Separate injection list and main poison list, to test out limits on
  injected poison list being smaller than the main list.
* Poison list overflow event (needs event log support in general)
* Connecting up to the poison list error record generation (rather complex
  and not needed for currently kernel handling testing).

As the kernel code is currently fairly simple, it is likely that the above
does not yet matter but who knows what will turn up in future!

Kernel patches:
 [PATCH v7 0/6] CXL Poison List Retrieval & Tracing
 [e-mail]
 [PATCH v2 0/6] cxl: CXL Inject & Clear Poison
 [e-mail]

Ira Weiny (2):
  hw/cxl: Introduce cxl_device_get_timestamp() utility function
  bswap: Add the ability to store to an unaligned 24 bit field

Jonathan Cameron (4):
  hw/cxl: rename mailbox return code type from ret_code to CXLRetCode
  hw/cxl: QMP based poison injection support
  hw/cxl: Add poison injection via the mailbox.
  hw/cxl: Add clear poison mailbox command support.

 hw/cxl/cxl-device-utils.c   |  15 ++
 hw/cxl/cxl-mailbox-utils.c  | 283 ++++++++++++++++++++++++++++++------
 hw/mem/cxl_type3.c          |  92 ++++++++++++
 hw/mem/cxl_type3_stubs.c    |   6 +
 include/hw/cxl/cxl_device.h |  23 +++
 include/qemu/bswap.h        |  23 +++
 qapi/cxl.json               |  18 +++
 7 files changed, 418 insertions(+), 42 deletions(-)

Comments

Philippe Mathieu-Daudé March 14, 2023, 6:32 a.m. UTC | #1

Hi Jonathan,

On 3/3/23 16:09, Jonathan Cameron wrote:
> Note there are several series ahead of this one and in particular
> the RAS error injection series needs some QAPI review.
> The QAPI stuff in this patch is similar but in essence very similar
> to what we have in that series.
> 
> Whilst I'm an always an optimist, this may well end up as 8.1 material
> now.


> Ira Weiny (2):
>    hw/cxl: Introduce cxl_device_get_timestamp() utility function
>    bswap: Add the ability to store to an unaligned 24 bit field
> 
> Jonathan Cameron (4):
>    hw/cxl: rename mailbox return code type from ret_code to CXLRetCode
>    hw/cxl: QMP based poison injection support
>    hw/cxl: Add poison injection via the mailbox.
>    hw/cxl: Add clear poison mailbox command support.
> 
>   hw/cxl/cxl-device-utils.c   |  15 ++
>   hw/cxl/cxl-mailbox-utils.c  | 283 ++++++++++++++++++++++++++++++------
>   hw/mem/cxl_type3.c          |  92 ++++++++++++
>   hw/mem/cxl_type3_stubs.c    |   6 +
>   include/hw/cxl/cxl_device.h |  23 +++

There is a '64' magic number used in various places, I haven't tried to
figure what is / where it comes from, but a CXL #definition for it could
make sense.