Request for review: JDK-8013041: guarantee(this->is8bit(imm8)) failed: Short forward jump exceeds 8-bit offset

[Jiangli Zhou]

Hi Dan,

Thanks for the review! Please see more comments below.

On 04/24/2013 02:55 PM, Daniel D. Daugherty wrote:
> On 4/24/13 1:37 PM, Jiangli Zhou wrote:
>> Please review the following fix for JDK-8013041 
>> <https://jbs.oracle.com/bugs/browse/JDK-8013041>. The 'jmpb' using 
>> 8-bit displacement needs to be replaced with 'jmp' with more 
>> instructions now generated between the jump instruction and the target.
>>
>> http://cr.openjdk.java.net/~jiangli/8013041/webrev.00/
>
> Thumbs up.
>
> src/cpu/x86/vm/templateInterpreter_x86_32.cpp
> src/cpu/x86/vm/templateInterpreter_x86_64.cpp
>     No comments on either file.
>
> Jiangli, any idea what caused the code size increase that made the 
> jump size
> more than what a 'jmpb' could handle? You might want to link to that 
> issue,
> if you happen to know what caused this.

The new code after the 'jmpb' for accessing the MethodCounters causes 
the target label to be further away from the original jump instruction. 
So the problem shows up when the code path is enabled.

>
> Also, will this change be pushed to Comp_Baseline so the compiler
> guys get access to fix more quickly?

Yes, that's the plan. Just got the approval from the compiler team.

Thanks!
Jiangli

>
> Dan
>
>
>>
>> Thanks,
>>
>> Jiangli
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.openjdk.java.net/pipermail/hotspot-runtime-dev/attachments/20130424/743f3f5c/attachment.html