From: Wei Hu <weh@linux.microsoft.com>
To: longli@kernel.org, kotaranov@microsoft.com, kuba@kernel.org,
davem@davemloft.net, pabeni@redhat.com, edumazet@google.com,
andrew+netdev@lunn.ch, jgg@ziepe.ca, leon@kernel.org,
haiyangz@microsoft.com, wei.liu@kernel.org, decui@microsoft.com,
shradhagupta@linux.microsoft.com, horms@kernel.org,
ernis@linux.microsoft.com, stephen@networkplumber.org
Cc: netdev@vger.kernel.org, linux-rdma@vger.kernel.org,
linux-hyperv@vger.kernel.org, linux-kernel@vger.kernel.org,
gargaditya@linux.microsoft.com, kees+treewide@kernel.org,
kees@kernel.org, mawasthi@linux.microsoft.com, weh@microsoft.com,
edumazet@kernel.org
Subject: [PATCH net-next v9 0/4] net: mana: concurrent HWC requests and dynamic queue depth
Date: Thu, 22 Oct 2026 15:02:58 +0000 [thread overview]
Message-ID: <cover.1792679388.git.weh@linux.microsoft.com> (raw)
This four-patch series adds concurrent HWC request ownership and
PF-directed HWC queue depth to MANA. It starts with depth-one bootstrap
queues and keeps setup private until the final channel is tested.
I am continuing Long Li's series. Original per-patch author identities,
dates and sign-offs are retained.
Changes in v9 (v8 -> v9):
- Rebase onto net-next 207dd12bc195153fe6ca614d8add9ecb0b636270.
This base already includes MANA BQL support (799cdeaa5c39).
- In patch 3, replace the independent admission deadline with
bitmap/waitqueue admission. Healthy contention waits for a free slot
instead of returning EBUSY before a destructive command is posted.
- Wake admission on slot release, recorded response abandonment,
cancellation and channel stop. With no available slot, abandonment
or cancellation returns ETIMEDOUT; a stopped channel returns ENODEV.
- Preserve available-slot best-effort posting after cancellation and
keep the response timeout for a posted command.
- Suppress expected ETIMEDOUT/ENODEV logging after cancellation or HWC
unpublication, while retaining live-channel and other diagnostics.
- Correct the eight reported source-line length warnings across all
four patches under netdev's explicit 80-column check.
Healthy admission no longer has a separate elapsed-time deadline,
matching the upstream waiting policy rather than treating healthy
contention as a channel failure. Response ownership, timed-out slot
quarantine and sender draining remain in place.
The PF specifies the HWC queue depth and corresponding ring sizes. It
reports the required depth through HWC_INIT_DATA_QUEUE_DEPTH, and the
VF rebuilds the rings to implement that PF requirement. Ring-sizing and
establishment behavior is unchanged from v8.
The pre-existing individual Ethernet/RDMA CQ publication and lifetime
issues, malformed-response RX reposting, and wider control-command DMA
failure policy remain separate work. This series does not claim to fix
them. Existing protections introduced later in the series are preserved
in their owning patches.
Validation on this revision:
- Clean application of all four patches to the net-next base above,
with no conflict resolution or three-way fallback.
- GCC and Clang W=1 builds of the MANA Ethernet/RDMA target objects at
the baseline and each of the four intermediate commits.
- Full-patch netdev-style checkpatch with --strict --codespell and
--max-line-length=80: no errors, warnings or CHECK diagnostics.
- Bounded actual-source C models cover admission and response
ownership, cancellation/logging, destructive callers, retained state,
initialization, rebuild and producer limits, with negative controls.
- Exact application and author/date/message checks on the outgoing
patch emails.
These checks are static/offline. The models use typed userspace
substitutes and modeled hardware acknowledgments; no new full-kernel,
hardware, performance or runtime qualification is claimed for v9.
This HWC series remains independent of queue-set replacement. The
separately prepared queue-set series is not included here; its dependency
and shared CQ-table changes must be reconciled separately.
v8:
https://lore.kernel.org/all/cover.1791998462.git.weh@linux.microsoft.com/
v7:
https://lore.kernel.org/all/cover.1791561995.git.weh@linux.microsoft.com/
v6:
https://lore.kernel.org/all/cover.1790665894.git.weh@linux.microsoft.com/
Long Li (4):
net: mana: prepare HWC ownership for safe reinitialization
net: mana: give each HWC message slot its own completion state
net: mana: support concurrent HWC requests
net: mana: add dynamic HWC queue depth with reinit path
drivers/infiniband/hw/mana/main.c | 19 +-
.../net/ethernet/microsoft/mana/gdma_cdx.c | 5 +-
.../net/ethernet/microsoft/mana/gdma_main.c | 158 ++-
.../net/ethernet/microsoft/mana/gdma_pci.c | 15 +-
.../net/ethernet/microsoft/mana/hw_channel.c | 1025 ++++++++++++++---
.../net/ethernet/microsoft/mana/shm_channel.c | 58 +-
include/net/mana/gdma.h | 17 +
include/net/mana/hw_channel.h | 70 +-
include/net/mana/shm_channel.h | 10 +-
9 files changed, 1155 insertions(+), 222 deletions(-)
base-commit: 207dd12bc195153fe6ca614d8add9ecb0b636270
--
2.43.0
next reply other threads:[~2026-10-22 15:03 UTC|newest]
Thread overview: 6+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-10-22 15:02 Wei Hu [this message]
2026-10-22 15:02 ` [PATCH net-next v9 1/4] net: mana: prepare HWC ownership for safe reinitialization Wei Hu
2026-10-22 15:03 ` [PATCH net-next v9 2/4] net: mana: give each HWC message slot its own completion state Wei Hu
2026-10-22 15:03 ` [PATCH net-next v9 3/4] net: mana: support concurrent HWC requests Wei Hu
2026-10-22 15:03 ` [PATCH net-next v9 4/4] net: mana: add dynamic HWC queue depth with reinit path Wei Hu
2026-10-22 15:23 ` [PATCH net-next v9 0/4] net: mana: concurrent HWC requests and dynamic queue depth Jakub Kicinski
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=cover.1792679388.git.weh@linux.microsoft.com \
--to=weh@linux.microsoft.com \
--cc=andrew+netdev@lunn.ch \
--cc=davem@davemloft.net \
--cc=decui@microsoft.com \
--cc=edumazet@google.com \
--cc=edumazet@kernel.org \
--cc=ernis@linux.microsoft.com \
--cc=gargaditya@linux.microsoft.com \
--cc=haiyangz@microsoft.com \
--cc=horms@kernel.org \
--cc=jgg@ziepe.ca \
--cc=kees+treewide@kernel.org \
--cc=kees@kernel.org \
--cc=kotaranov@microsoft.com \
--cc=kuba@kernel.org \
--cc=leon@kernel.org \
--cc=linux-hyperv@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-rdma@vger.kernel.org \
--cc=longli@kernel.org \
--cc=mawasthi@linux.microsoft.com \
--cc=netdev@vger.kernel.org \
--cc=pabeni@redhat.com \
--cc=shradhagupta@linux.microsoft.com \
--cc=stephen@networkplumber.org \
--cc=weh@microsoft.com \
--cc=wei.liu@kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
all inboxes | Powered by JetHome®