On 10.02.2022 18:25, Peter Müller wrote:
> Hello Matthias,

Hi Peter,

> hello *,
> 
> by coincidence, I just stumbled across a Squid 5.4 source tarball on source.ipfire.org:

Yep. ;-)

>> pmueller(a)people01:~$ ls -lah /pub/sources/source-2.x/squid-5.*
>> -rw-r--r-- 1 mfischer people 2.5M Aug  2  2021 /pub/sources/source-2.x/squid-5.1.tar.xz
>> -rw-r--r-- 1 mlorenz  people 2.5M Oct  6 07:44 /pub/sources/source-2.x/squid-5.2.tar.xz
>> -rw-r--r-- 1 mfischer people 2.5M Dec  7 18:42 /pub/sources/source-2.x/squid-5.3.tar.xz
>> -rw-r--r-- 1 mfischer people 2.5M Feb  7 17:10 /pub/sources/source-2.x/squid-5.4.tar.xz
> 
> I therefore presume you are working on this one. :-)

You're right - I do.

> While Squid 5.2, as it is now present in the upcoming Core Update 164, works fine, I
> observed at least one memory leak in it that eventually caused swapping on a productive
> infrastructure (under high load). Squid 5.4 should have that fixed.

I hope so, but I wanted to be sure before pushing.

> Because of this, I'd like to have Squid 5.4 in Core Update 164, so our users won't have
> to suffer from this bug. It is good to see you are taking care of it. Please do not feel
> pressured in any way, I just thought I mention this to you. :-)

No problem - one of my "Devels" is already busy... ;-)

While I'm writing this he is completing the final build. Looks good.

I nearly pushed 5.4 yesterday but then I saw this patch
(http://www.squid-cache.org/Versions/v5/changesets/):

http://www.squid-cache.org/Versions/v5/changesets/squid-5-1332f8d606485b5a2f57277634c2f6f7855bd9a6.patch

=> Bug #5055: FATAL FwdState::noteDestinationsEnd exception: opening (#967)

And now I'm not sure: shall I include this patch or not!?

Its a huge one.

Excerpt:

"Bug 5055: FATAL FwdState::noteDestinationsEnd exception: opening (#967)

    * Bug 5055: FATAL FwdState::noteDestinationsEnd exception: opening

    The bug was caused by commit 25b0ce4. Other known symptoms are:

        assertion failed: store.cc:1793: "isEmpty()"
        assertion failed: FwdState.cc:501: "serverConnection() == conn"
        assertion failed: FwdState.cc:1037: "!opening()"

    This change has several overlapping parts. Unfortunately, merging
    individual parts is both difficult and likely to cause crashes.
..."

What do you think?

> Thanks in advance for your efforts, and best regards,

You're welcome!

> Peter Müller

Best,
Matthias