mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Avi Weiss <thnkslprpt@gmail.com>
To: Jon Mason <jdmason@kudzu.us>
Cc: Dave Jiang <dave.jiang@intel.com>,
	Allen Hubbe <allenbh@gmail.com>, Koichiro Den <den@valinux.co.jp>,
	Frank Li <Frank.Li@nxp.com>,
	ntb@lists.linux.dev, linux-kernel@vger.kernel.org,
	Avi Weiss <thnkslprpt@gmail.com>
Subject: [PATCH v2 0/2] NTB: epf: serialize MW setup and report reliable errors
Date: Sun,  1 Nov 2026 15:57:04 +0200	[thread overview]
Message-ID: <cover.1793540847.git.thnkslprpt@gmail.com> (raw)

This series fixes two related NTB EPF memory-window issues.

The first patch serializes shared MW parameter writes with command
submission, preventing concurrent setups from using each other's
DMA addresses.

The second patch propagates definite endpoint command errors while
completion status can still be associated with the current command.

Changes since v1:

- koichiro pointed out that an endpoint may complete CONFIGURE_MW after
the host times out. Propagating -ETIMEDOUT can cause clients to free
DMA memory while the endpoint is still configuring the mapping.

Returning success only for the timed-out command avoids that immediate
regression, but was insufficient in the end b/c a late error from 
command A may be observed during command B. Returning that stale error
can cause B's buffer to be freed before B completes.

The second patch records whether any command has timed out and
preserves the original MW return behavior thereafter.

The first patch also prevents a pre-timeout regression: without
parameter serialization, propagating an explicit rejection from one
setup can release memory referenced by another concurrent setup.

v1: https://lists.openwall.net/linux-kernel/2026/10/27/2144
Review: https://lists.openwall.net/linux-kernel/2026/10/28/466

Avi Weiss (2):
  NTB: epf: serialize memory window parameter writes
  NTB: epf: propagate MW command errors while status is reliable

 drivers/ntb/hw/epf/ntb_hw_epf.c | 37 ++++++++++++++++++++++++++-------
 1 file changed, 29 insertions(+), 8 deletions(-)


base-commit: cbbddb6b08cc9900a87cccaa9b22bbe06aa4f50a
-- 
2.43.0


             reply	other threads:[~2026-11-01 13:57 UTC|newest]

Thread overview: 5+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-11-01 13:57 Avi Weiss [this message]
2026-11-01 13:57 ` [PATCH v2 1/2] NTB: epf: serialize memory window parameter writes Avi Weiss
2026-11-01 15:53   ` Koichiro Den
2026-11-01 13:57 ` [PATCH v2 2/2] NTB: epf: propagate MW command errors while status is reliable Avi Weiss
2026-11-01 15:56   ` Koichiro Den

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.1793540847.git.thnkslprpt@gmail.com \
    --to=thnkslprpt@gmail.com \
    --cc=Frank.Li@nxp.com \
    --cc=allenbh@gmail.com \
    --cc=dave.jiang@intel.com \
    --cc=den@valinux.co.jp \
    --cc=jdmason@kudzu.us \
    --cc=linux-kernel@vger.kernel.org \
    --cc=ntb@lists.linux.dev \
    /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®