review for 7043301: assert(locals < caller->fp() || locals > (caller->fp() + 16)) failed: locals in save area (original) (raw)

Vladimir Kozlov vladimir.kozlov at oracle.com
Mon May 9 18:52:35 PDT 2011

• Previous message: review for 7043301: assert(locals < caller->fp() || locals > (caller->fp() + 16)) failed: locals in save area
• Next message: review for 7043301: assert(locals < caller->fp() || locals > (caller->fp() + 16)) failed: locals in save area
• Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

Thank you, Tom

Vladimir

Tom Rodriguez wrote:

They all run fine too.

tom On May 9, 2011, at 4:55 PM, Vladimir Kozlov wrote:

Tom,

It looks good. Could you test the fix with failed tests from last Nightly (-XX:+DeoptimizeALot)? Thanks, Vladimir Tom Rodriguez wrote: http://cr.openjdk.java.net/~never/7043301 7043301: assert(locals < caller->fp() || locals > (caller->fp() + 16)) failed: locals in save area Reviewed-by: During the fix for 7009361 I mistakenly simplified the computedspadjustment by replacing it with roundedcls but one is for the caller and the other is for the callee. This could result in some rare cases where the frame shrunk to be smaller than it should have been. The fix is to restore the original computation of this value. Tested with failing test case. I also corrected the frame validation code which should have complained about this.

• Previous message: review for 7043301: assert(locals < caller->fp() || locals > (caller->fp() + 16)) failed: locals in save area
• Next message: review for 7043301: assert(locals < caller->fp() || locals > (caller->fp() + 16)) failed: locals in save area
• Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

More information about the hotspot-compiler-dev mailing list