public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
* [PATCH] docs: bug-bisect: Add a note about bisecting -next
@ 2024-10-22 15:59 Mark Brown
  2024-10-23 18:05 ` Thorsten Leemhuis
  0 siblings, 1 reply; 3+ messages in thread
From: Mark Brown @ 2024-10-22 15:59 UTC (permalink / raw)
  To: Thorsten Leemhuis, Jonathan Corbet; +Cc: linux-doc, linux-kernel, Mark Brown

We don't explicitly mention anywhere in the kernel tree that bisects
between -next versions won't work well and it's better to bisect between
mainline and -next. Let's add a note about that to try to help people avoid
this particular gotcha.

Signed-off-by: Mark Brown <broonie@kernel.org>
---
 Documentation/admin-guide/bug-bisect.rst | 12 ++++++++++++
 1 file changed, 12 insertions(+)

diff --git a/Documentation/admin-guide/bug-bisect.rst b/Documentation/admin-guide/bug-bisect.rst
index 585630d14581c7e0bdf9dd3b66d427793d41925b..eef6921a9542ef276c097e5861ca4efe5812ea0d 100644
--- a/Documentation/admin-guide/bug-bisect.rst
+++ b/Documentation/admin-guide/bug-bisect.rst
@@ -109,6 +109,18 @@ With that the process is complete. Now report the regression as described by
 Documentation/admin-guide/reporting-issues.rst.
 
 
+Bisecting linux-next
+--------------------
+
+Since linux-next is a series of merges rebuilt every day starting from
+Linus' tree there is no commmon history between multiple versions of
+-next. This means that the history of a given -next release won't
+include prior -next releases which confuses bisect if you try to
+bisect between them. Bisects will run much better if performed between
+-next and the commit in Linus' tree which that version of -next is
+based on instead.
+
+
 Additional reading material
 ---------------------------
 

---
base-commit: 8e929cb546ee42c9a61d24fae60605e9e3192354
change-id: 20241022-doc-bisect-next-d47c6ace8a95

Best regards,
-- 
Mark Brown <broonie@kernel.org>


^ permalink raw reply related	[flat|nested] 3+ messages in thread

* Re: [PATCH] docs: bug-bisect: Add a note about bisecting -next
  2024-10-22 15:59 [PATCH] docs: bug-bisect: Add a note about bisecting -next Mark Brown
@ 2024-10-23 18:05 ` Thorsten Leemhuis
  2024-10-23 18:14   ` Mark Brown
  0 siblings, 1 reply; 3+ messages in thread
From: Thorsten Leemhuis @ 2024-10-23 18:05 UTC (permalink / raw)
  To: Mark Brown, Jonathan Corbet; +Cc: linux-doc, linux-kernel

On 22.10.24 17:59, Mark Brown wrote:
> We don't explicitly mention anywhere in the kernel tree that bisects
> between -next versions won't work well and it's better to bisect between
> mainline and -next. Let's add a note about that to try to help people avoid
> this particular gotcha.
> 
> Signed-off-by: Mark Brown <broonie@kernel.org>

Thx for this!

> ---
>  Documentation/admin-guide/bug-bisect.rst | 12 ++++++++++++
>  1 file changed, 12 insertions(+)
> [...]
> +Bisecting linux-next
> +--------------------
> +
> +Since linux-next is a series of merges rebuilt every day starting from
> +Linus' tree there is no commmon history between multiple versions of
> +-next. This means that the history of a given -next release won't
> +include prior -next releases which confuses bisect if you try to
> +bisect between them. Bisects will run much better if performed between
> +-next and the commit in Linus' tree which that version of -next is
> +based on instead.

Hmmmm. That describes the problem (somewhat in style of a "reference
guide"), but does not really match the approach the rest of the text
uses (which is more of an "howto"). So for me it feels like a bad fit,
even if users of linux-next are less likely to need a howto.

How about something like this instead:

---

Bisecting linux-next
--------------------

If you face a problem that only happens in linux-next, bisect between
the linux-next branches "stable" and "master". Use these commands to
start the bisection for a linux-next tree you added as a remote called
"next"::

  git bisect start
  git bisect good next/stable
  git bisect bad next/master

The "stable" branch refers to the state of linux-mainline the current
linux-next release is based on -- so it should be free of any problems
that show up in -next, but not in Linus' tree.

This approach will mean that you have to bisect across a wide range of
changes, some of which you might have used in earlier linux-next
releases without problems. Sadly it's impossible to bisect from one
linux-next release to a later one (say between next-20241022 and
next-20241023), as the -next specific changes of the two share no common
history.

---

I did not check the commands in that text yet, hope I did not do
anything stupid there.

Ciao, Thorsten

^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: [PATCH] docs: bug-bisect: Add a note about bisecting -next
  2024-10-23 18:05 ` Thorsten Leemhuis
@ 2024-10-23 18:14   ` Mark Brown
  0 siblings, 0 replies; 3+ messages in thread
From: Mark Brown @ 2024-10-23 18:14 UTC (permalink / raw)
  To: Thorsten Leemhuis; +Cc: Jonathan Corbet, linux-doc, linux-kernel

[-- Attachment #1: Type: text/plain, Size: 484 bytes --]

On Wed, Oct 23, 2024 at 08:05:33PM +0200, Thorsten Leemhuis wrote:

> How about something like this instead:
> 
> ---
> 
> Bisecting linux-next
> --------------------
> 
> If you face a problem that only happens in linux-next, bisect between
> the linux-next branches "stable" and "master". Use these commands to
> start the bisection for a linux-next tree you added as a remote called
> "next"::

This looks good to me.

Reviewed-by: Mark Brown <broonie@kernel.org>

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]

^ permalink raw reply	[flat|nested] 3+ messages in thread

end of thread, other threads:[~2024-10-23 18:14 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2024-10-22 15:59 [PATCH] docs: bug-bisect: Add a note about bisecting -next Mark Brown
2024-10-23 18:05 ` Thorsten Leemhuis
2024-10-23 18:14   ` Mark Brown

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox