On 02/04/11 03:37, Alex Rousskov wrote:
> On 03/31/2011 04:59 PM, Amos Jeffries wrote:
>> On 01/04/11 09:36, Kinkie wrote:
>>>> I left the "if (true)" guard to prevent indentation changes to the
>>>> pretty long logRequest method, making the patch more difficult to
>>>> understand. I will remove that otherwise useless guard during commit
>>>> unless you think it should stay.
>>>
>>> +1, with the suggestion of possibly making the debugs() statement have
>>> a lower priority (that is, make it harder to print).
>>
>> + and ditto on the debugs. 5 maybe 6for that IMO.
>>
>> And yes, please remove the if (true) and run the
>> scripts/sourcemaintenance.sh over it before commit.
>
> I used level-2 debugging in case this change leads to logging coredumps:
> Amos used that approach in a different context recently, arguing that it
> would be easier to request/get this debugging information that way.
>
> Formatted and committed to trunk with level 5 as r11337.
>
Are you referring to:
http://www.squid-cache.org/Versions/v3/3.1/changesets/squid-3.1-10245.patch
... ow-level debug raised to important after the case was confirmed to
hit some users badly as core dumps. rev10245 added after it proved to
be extremely difficult to find the dump() details manually.
A bad URL seems easy to detect with the number of times URL gets
displayed. If you know otherwise feel free to correct us and use
appropriate levels.
Amos
-- Please be using Current Stable Squid 2.7.STABLE9 or 3.1.11 Beta testers wanted for 3.2.0.5Received on Sun Apr 03 2011 - 06:05:35 MDT
This archive was generated by hypermail 2.2.0 : Sun Apr 03 2011 - 12:00:10 MDT