On 10.11.21 20:16, Alexei Starovoitov wrote:
> On Wed, Nov 10, 2021 at 10:01 AM Thorsten Leemhuis
> <regressions@xxxxxxxxxxxxx> wrote:
>> On 10.11.21 17:50, Alexei Starovoitov wrote:
>>> On Wed, Nov 10, 2021 at 11:41:09AM +0000, Lorenz Bauer wrote:
>>>>
>>>> uid changes on every invocation, and therefore regsafe() returns false?
>>>
>>> That's correct.
>>> Could you please try the following fix.
>>> I think it's less risky and more accurate than what I've tried before.
>>>
>>> >From be7736582945b56e88d385ddd4a05e13e4bc6784 Mon Sep 17 00:00:00 2001
>>> From: Alexei Starovoitov <ast@xxxxxxxxxx>
>>> Date: Wed, 10 Nov 2021 08:47:52 -0800
>>> Subject: [PATCH] bpf: Fix inner map state pruning regression.
>>>
>>> Fixes: 3e8ce29850f1 ("bpf: Prevent pointer mismatch in bpf_timer_init.")
>>> Signed-off-by: Alexei Starovoitov <ast@xxxxxxxxxx>
>>
>> Thanks for working on a fix for this regression. There is one small
>> detail that afaics could be improved (maybe you left that for later, if
>> that's the case please simply stop reading and ignore this mail):
>>
>> The commit message would benefit from a link to the regression report.
>> This is explained in Documentation/process/submitting-patches.rst, which
>> recently was changed slightly to make this aspect clearer:
>> https://git.kernel.org/linus/1f57bd42b77c
>
> I don't think you're familiar with bpf process of applying patches.
That's totally correct, but why should someone working as a regression
tracker for all of the Linux kernel has to know the exact inner workings
of all the various subsystems?
> Please take a look at bpf tree. The 'Link:' exists for every commit.
One of us is missing something. I might be wrong, but you afaics mean
that every commit has a 'Link:' to the latest review posting added when
a patch is committed. Many thx for that, I really like that, as it made
my life easier a few times in the past already.
But that was not what I was asking got. I was asking for a 'Link:' to
the *regression report*. To quote myself again from above:
>> The commit message would benefit from a link to the regression report.
>> [...] https://git.kernel.org/linus/1f57bd42b77c
To quote from there:
```
In case your patch fixes a bug, for example, add a tag with a URL
referencing the report in the mailing list archives or a bug tracker
````
Or what am I missing? Ciao, Thorsten
