mbox series

[GIT,PULL] Compute Express Link (CXL) Fixes for 6.10-rc7

Message ID e5477d05-93e0-4268-90ca-581e3f492b4e@intel.com
State Not Applicable
Headers show
Series [GIT,PULL] Compute Express Link (CXL) Fixes for 6.10-rc7 | expand

Pull-request

git://git.kernel.org/pub/scm/linux/kernel/git/cxl/cxl.git tags/cxl-fixes-6.10-rc7

Message

Dave Jiang July 1, 2024, 5:53 p.m. UTC
Hi Linus, please pull from

  git://git.kernel.org/pub/scm/linux/kernel/git/cxl/cxl.git tags/cxl-fixes-6.10-rc7

...to receive several fixes for the CXL subsystem.

A fix to deal with NULL cxl_nvd pointer during persistent memory auto assembly.

A fix to address NULL pointer dereference in CXL region lookup.

A fix to add proper checking for CXL region interleave capability during assembly.

A kdoc fix for CXL documentation.

This has build success notification from kbuild-robot. It has been in the -next
for several days with no reported issues.

---

The following changes since commit 6ba59ff4227927d3a8530fc2973b80e94b54d58f:

  Linux 6.10-rc4 (2024-06-16 13:40:16 -0700)

are available in the Git repository at:

  git://git.kernel.org/pub/scm/linux/kernel/git/cxl/cxl.git tags/cxl-fixes-6.10-rc7

for you to fetch changes up to a0f39d51dbf72c28283bd201b97559ed82bc0fe5:

  cxl: documentation: add missing files to cxl driver-api (2024-06-25 14:45:27 -0700)

----------------------------------------------------------------
cxl fixes for v6.10-rc7

- Fix no cxl_nvd during pmem region auto-assemble
- Avoid NULLL pointer dereference in region lookup
- Add missing checks to interleave capability
- Add cxl kdoc fix to address document compilation error

----------------------------------------------------------------
Alison Schofield (1):
      cxl/region: Avoid null pointer dereference in region lookup

Li Ming (1):
      cxl/mem: Fix no cxl_nvd during pmem region auto-assembling

Yao Xingtao (2):
      cxl/region: check interleave capability
      cxl: documentation: add missing files to cxl driver-api

 Documentation/driver-api/cxl/memory-devices.rst |  15 ++++
 drivers/cxl/core/hdm.c                          |  13 +++
 drivers/cxl/core/pmem.c                         |  16 ++--
 drivers/cxl/core/region.c                       | 103 ++++++++++++++++++++++--
 drivers/cxl/cxl.h                               |   6 +-
 drivers/cxl/cxlmem.h                            |  21 +++--
 drivers/cxl/mem.c                               |  17 ++--
 tools/testing/cxl/test/cxl.c                    |   4 +
 8 files changed, 170 insertions(+), 25 deletions(-)

Comments

Linus Torvalds July 1, 2024, 8:12 p.m. UTC | #1
On Mon, 1 Jul 2024 at 10:53, Dave Jiang <dave.jiang@intel.com> wrote:
>
> A fix to add proper checking for CXL region interleave capability during assembly.

I pulled this, but this should not have been a post-rc6 "fix".

That code is absolutely new development. Any interleave mis-match has
never worked in the past afaik, and this really looks like a new
feature to me.

So I think this should have been a "next merge window development"
thing. Fixes are for things that have broken or cause acute problems.
Not for "this never worked and nobody even reported it".

             Linus
pr-tracker-bot@kernel.org July 1, 2024, 8:16 p.m. UTC | #2
The pull request you sent on Mon, 1 Jul 2024 10:53:44 -0700:

> git://git.kernel.org/pub/scm/linux/kernel/git/cxl/cxl.git tags/cxl-fixes-6.10-rc7

has been merged into torvalds/linux.git:
https://git.kernel.org/torvalds/c/73e931504f8e0d42978bfcda37b323dbbd1afc08

Thank you!
Dave Jiang July 1, 2024, 9 p.m. UTC | #3
On 7/1/24 1:12 PM, Linus Torvalds wrote:
> On Mon, 1 Jul 2024 at 10:53, Dave Jiang <dave.jiang@intel.com> wrote:
>>
>> A fix to add proper checking for CXL region interleave capability during assembly.
> 
> I pulled this, but this should not have been a post-rc6 "fix".

Thank you
> 
> That code is absolutely new development. Any interleave mis-match has
> never worked in the past afaik, and this really looks like a new
> feature to me.
> 
> So I think this should have been a "next merge window development"
> thing. Fixes are for things that have broken or cause acute problems.
> Not for "this never worked and nobody even reported it".

I contemplated whether this is a fix or should be left to the next merge window. Next time in doubt, I'll leave it to the next merge window. Sorry about that. 

> 
>              Linus