all lists on lists.proxmox.com
 help / color / mirror / Atom feed
From: 結友 <miagetegorann@gmail.com>
To: t.lamprecht@proxmox.com
Cc: pve-devel@lists.proxmox.com
Subject: Re: [PATCH pve-kernel v4 1/2] d/scripts: correctly compare namespaced ABI exports
Date: Tue, 17 Nov 2026 08:50:15 +0900	[thread overview]
Message-ID: <20261116235015.31-1-miagetegorann@gmail.com> (raw)
In-Reply-To: <aa154653-9aa2-4554-be6a-d14253e34237@proxmox.com>

Thanks for the review!

> why python tests and not perl, like the script doing the abi export is using? Could be
> probably also a simpler array of input, expected tuples to iterate over, but not a must.

No good reason. I dropped the tests in v5, see below.

> Also, these tests don't really give us any coverage for the real ABI output? if that
> diverges (again) we still just miss that? Some actual check and error on unexpected
> input above would be probably more helpful.

Agreed, the tests are not run by the build, so they would not notice
that. The check itself already does though: parse_symbol() dies for
every entry matching neither of the two layouts, and the callers die on
duplicate symbols. As abicheck runs on the freshly generated file on
each build, a diverging format fails the build, e.g.:

  EE: abi-7.0.14-29-pve-amd64:32728: malformed ABI entry: 'garbage'

That is also the case with SKIPABI=1 and in the no-fail mode after an
ABI bump, as the comparison cannot be trusted then. If you would rather
have it be a warning in those modes, I can change that.

While checking this again I noticed that the current parsing is also
wrong for more than the VMBus move: for the 2557 namespaced entries it
takes the namespace as export type, so a namespaced export changing
from EXPORT_SYMBOL to EXPORT_SYMBOL_GPL is not detected. I added that
to the commit message.

The two layouts come from awk skipping the empty namespace column in
abi-generate. I kept reading both in abi-check to avoid regenerating
the ABI file, but can instead make abi-generate emit a single layout
if you prefer that.

I just sent a v5:
"[PATCH pve-kernel v5 0/2] fix #8027: restore built-in Hyper-V VMBus"



  reply	other threads:[~2026-11-16 23:50 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-10-24  5:24 [PATCH pve-kernel v4 0/2] fix #8027: restore built-in Hyper-V VMBus 結友
2026-10-24  5:24 ` [PATCH pve-kernel v4 1/2] d/scripts: correctly compare namespaced ABI exports 結友
2026-11-16 21:48   ` Thomas Lamprecht
2026-11-16 23:50     ` 結友 [this message]
2026-10-24  5:24 ` [PATCH pve-kernel v4 2/2] fix #8027: config: enable CONFIG_HYPERV_VMBUS 結友
2026-11-16 23:54   ` Thomas Lamprecht
2026-11-17  0:03     ` 結友
2026-11-09 18:39 ` [PATCH pve-kernel v4 0/2] fix #8027: restore built-in Hyper-V VMBus 山﨑結友

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=20261116235015.31-1-miagetegorann@gmail.com \
    --to=miagetegorann@gmail.com \
    --cc=pve-devel@lists.proxmox.com \
    --cc=t.lamprecht@proxmox.com \
    /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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.
Service provided by Proxmox Server Solutions GmbH | Privacy | Legal