RFR(M): : Various minor code improvements (compiler)

Tobias Hartmann tobias.hartmann at oracle.com
Mon May 9 07:36:32 UTC 2016


Hi Goetz,

On 06.05.2016 10:01, Lindenmaier, Goetz wrote:
> Please review this change. I please need a sponsor.
> 
> http://cr.openjdk.java.net/~goetz/wr16/8140594-covCo/webrev.00/index.html

This looks good to me. Thanks for the detailed explanation, please see some minor suggestions below.

I can sponsor your fix but since we are close to JDK 9 FC, we are not allowed to push enhancements to hs-comp. Shouldn't this be a "bug" anyway?

> os_linux.cpp
> 
> - Buffer not null terminated
> 
> - size of pointer passed to jio_snprintf()

Please also fix the indentation in line 6004.

> classLoader.cpp
> 
> - jio_snprintf does null termination. But it might return -1 if truncated,
> 
>   in this case array access at -1.

You can remove "int n;"

> generateOopMap.cpp
> 
> - Remaining fields not initialized.

I would put the loop bodies in a new line.

> relocator.cpp
> 
> - delta might be -4 ... assert returns.

I don't understand this change. If delta is -4, the assert in the baseline version is triggered. With your fix, the assert is triggered as well.

Best regards,
Tobias


More information about the hotspot-compiler-dev mailing list