[webkit-dev] Always-on diagnostic code Re: [webkit-changes] [96819] trunk/Source/WebCore

[Dan Bernstein]

On Oct 6, 2011, at 10:11 AM, Gavin Peters (蓋文彼德斯) wrote:

> Dan,
> 
> You're right.  I'm adding #if PLATFORM(CHROMIUM) to this change shortly.

Thank you.

> 
> - Gavin
> 
> On 6 October 2011 13:07, Dan Bernstein <mitz at apple.com> wrote:
> 
> On Oct 6, 2011, at 9:40 AM, gavinp at chromium.org wrote:
> 
>> Modified: trunk/Source/WebCore/dom/ScriptElement.h (96818 => 96819)
>> 
>> 
>> --- trunk/Source/WebCore/dom/ScriptElement.h	2011-10-06 16:37:35 UTC (rev 96818)
>> +++ trunk/Source/WebCore/dom/ScriptElement.h	2011-10-06 16:40:47 UTC (rev 96819)
>> @@ -113,6 +113,14 @@
>>        ZeroedInStopLoadRequest,
>>        ZeroedInNotifyFinished,
>>      } m_cachedScriptState;
>> +
>> +    // We grab a backtrace when we zero m_cachedScript, so that at later crashes
>> +    // we'll have a debuggable stack.
>> +    enum {
>> +        MaxBacktraceSize = 32
>> +    };
>> +    int m_backtraceSize;
>> +    void* m_backtrace[MaxBacktraceSize];
>>  };
> 
> This appears to increase the size of each ScriptElement instance by 256 bytes. I don’t know how bad a performance hit this is in real-world use, but it is most certainly not something all vendors would like to include in their releases. The way this change was made, however, it is almost inevitable that a vendor would end up unknowingly shipping this performance regression. This change was made on trunk, it is unconditionally compiled in, and there is nothing obvious tracking undoing this change.
> 
> I think this is the wrong way to incorporate diagnostic code into WebKit.
> 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.webkit.org/pipermail/webkit-dev/attachments/20111006/65dc0f07/attachment.html>