8220501: Improve c1_ValueStack locks handling

Claes Redestad claes.redestad at oracle.com
Tue Mar 12 15:06:35 UTC 2019


Hi Goetz,

yes, I've filed https://bugs.openjdk.java.net/browse/JDK-8220515 and
sent out a RFR for a partial removal.

/Claes

On 2019-03-12 15:39, Lindenmaier, Goetz wrote:
> Hi Claes,
> 
> it seems that your change breaks the slowdebug build:
> 
> .../src/hotspot/share/c1/c1_CFGPrinter.cpp: In member function ‘void CFGPrinterOutput::print_state(BlockBegin*)’:
> .../src/hotspot/share/c1/c1_CFGPrinter.cpp:169:7: error: ‘for_each_lock_value’ was not declared in this scope
>         for_each_lock_value(state, index, value) {
>         ^~~~~~~~~~~~~~~~~~~
> 
> Best regards,
>    Goetz.
> 
> 
> 
>> -----Original Message-----
>> From: hotspot-compiler-dev <hotspot-compiler-dev-
>> bounces at openjdk.java.net> On Behalf Of Claes Redestad
>> Sent: Dienstag, 12. März 2019 13:53
>> To: hotspot compiler <hotspot-compiler-dev at openjdk.java.net>
>> Subject: RFR: 8220501: Improve c1_ValueStack locks handling
>>
>> Hi,
>>
>> the _locks Values in ValueStack is often empty/unused, and allocating it
>> lazily is a small startup/warmup optimization.
>>
>> Webrev: http://cr.openjdk.java.net/~redestad/8220501/open.00/
>> Bug:    https://bugs.openjdk.java.net/browse/JDK-8220501
>>
>> Testing: tier1-3
>>
>> Thanks!
>>
>> /Claes


More information about the hotspot-compiler-dev mailing list