linux-kernel - Re: Linux 6.1-rc3

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20221101170015.GA1314742@roeck-us.net>
Date:   Tue, 1 Nov 2022 10:00:15 -0700
From:   Guenter Roeck <linux@...ck-us.net>
To:     Linus Torvalds <torvalds@...ux-foundation.org>
Cc:     Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
        Miguel Ojeda <ojeda@...nel.org>
Subject: Re: Linux 6.1-rc3

On Sun, Oct 30, 2022 at 03:35:51PM -0700, Linus Torvalds wrote:
> It's Sunday afternoon, so it must be time for an rc release.
> 
> I know I said last week that rc2 was unusually large. It turns out
> that rc3 is almost exactly the same size. But at least for an rc3
> release, that bigger size is a bit more normal: this is when people
> are starting to find problems and send in fixes for them.
> 
> So while rc2 was just _way_ bigger than usual, rc3 is only a bit
> larger than an average rc3 release is. But it's still on the largish
> side. I hope that things start calming down, and we'll start seeing
> the size of these rc's shrink. Please?
> 
> Unlike rc2, there's no one single reason for the bulk of the rc3
> changes. They're pretty much all over, with the usual distribution -
> drivers dominating (networking, gpu and sound are most noticeable, but
> there's a little bit of everything).
> 
> Outside of drivers, tool updates stand out, with selftests, perf, and
> the pm-graph tool all seeing a fair amount of changes.
> 
> And then we have the usual things: architecture updates, some
> filesystem work, and core kernel fixes (mainly networking and mm).
> 
> Anyway, while it isn't small, nothing looks particularly worrisome or
> strange, and I thin kyou can just scan the appended shortlog to get a
> feel for the kinds of fixes we have here. Please do give it more
> testing, and here's to hoping we'll start seeing the rc's shrink from
> now on.
> 

Build results:
	total: 152 pass: 152 fail: 0
Qemu test results:
	total: 500 pass: 500 fail: 0

No runtime warnings.

There is one spurious build error. It does not happen all the time,
and even on the same SHA it is not easy to reproduce. If I see it,
I may see it on one server reliably but not at all on another with
almost the same hardware configuration.

Building powerpc:allnoconfig ... failed
--------------
Error log:
Inconsistent kallsyms data
Try make KALLSYMS_EXTRA_PASS=1 as a workaround

I first noticed the error after commit 73bbb94466fd ("kallsyms: support
"big" kernel symbols"). Reverting this commit fixes the problem if and
when it is reproducible. I can not reproduce it with v6.1-rc3.
KALLSYMS_EXTRA_PASS=1 does _not_ fix the problem. Adding more kallsyms
passes (I tried up to 20) does not help either; it never stabilizes.
Instead, the data oscillates between runs (i.e., if it happens, even
runs produce the same kallsyms data, and odd runs produce the same data).

Maybe something changed between commit 73bbb94466fd and v6.1-rc3 and
the problem is gone for good. If not I'll report if I see it again.

Guenter