| Fri, 04 Oct 2013 12:45:57 +0100 |
adinn |
Merge
|
| Fri, 04 Oct 2013 12:44:40 +0100 |
adinn |
fix compare long against immediate to pass case for 0 into correct branch
|
| Tue, 01 Oct 2013 18:09:21 +0100 |
aph |
Remove a little bogosity
|
| Tue, 01 Oct 2013 16:17:58 +0100 |
aph |
JVMTI earlyreturn support
|
| Tue, 01 Oct 2013 15:01:30 +0100 |
aph |
Merge
|
| Mon, 30 Sep 2013 18:39:26 +0100 |
aph |
Assertion check only
|
| Mon, 30 Sep 2013 18:13:26 +0100 |
aph |
Fix register corrption caused by check_klass_subtype_slow_path.
|
| Mon, 30 Sep 2013 16:16:13 +0100 |
aph |
Rewrite checkcasting array copy stub
|
| Mon, 30 Sep 2013 08:55:28 +0100 |
Edward Nevill edward nevill |
Remove some debugging code
|
| Fri, 27 Sep 2013 18:04:00 +0100 |
aph |
Missing file from jdk8-b90 update
|
| Fri, 27 Sep 2013 15:48:05 +0100 |
Edward Nevill edward nevill |
Update aarch64 specific code to make it work with jdk8-b90
|
| Fri, 27 Sep 2013 15:44:21 +0100 |
Edward Nevill edward nevill |
Merge up to jdk8-b90
|
| Fri, 27 Sep 2013 12:34:47 +0100 |
adinn |
added testP/N to compare against object NULL directly against zr
|
| Tue, 24 Sep 2013 18:06:44 +0100 |
adinn |
make C2 runtime calls to generated code branch using bl rather than blrt
|
| Fri, 20 Sep 2013 15:59:37 +0100 |
Edward Nevill edward nevill |
Make zero cross compile build work again
|
| Fri, 20 Sep 2013 08:39:00 +0100 |
Edward Nevill edward nevill |
Added changes to aarch64 code for merge up to jdk8-b85
|
| Fri, 20 Sep 2013 08:36:25 +0100 |
Edward Nevill edward nevill |
Merged hotspot up to jdk8-b85 (Hotspot build hs25-b26)
|
| Thu, 19 Sep 2013 18:19:30 +0100 |
aph |
Merge
|
| Thu, 19 Sep 2013 18:07:10 +0100 |
aph |
Assert that we must not use the native copy routines with compressed oops.
|
| Thu, 19 Sep 2013 18:06:03 +0100 |
aph |
Fix off-by-one error in arraycopy barrier calculation.
|
| Thu, 19 Sep 2013 17:16:43 +0100 |
aph |
Implement generate_checkcast_copy()
|
| Thu, 19 Sep 2013 17:14:14 +0100 |
aph |
Merge
|
| Thu, 19 Sep 2013 17:03:09 +0100 |
aph |
Merge
|
| Thu, 19 Sep 2013 11:18:32 +0100 |
aph |
Arraycopy intrinsics
|
| Wed, 18 Sep 2013 11:19:41 +0100 |
aph |
Rough draft of arraycopy prototype
|
| Tue, 17 Sep 2013 17:27:50 +0100 |
adinn |
added a patchable nop at the start of a C2 method for use at deopt
|
| Tue, 17 Sep 2013 16:44:26 +0100 |
adinn |
Merge
|
| Tue, 17 Sep 2013 11:38:11 +0100 |
adinn |
initial unroll info should use stored fp as per x86
|
| Fri, 13 Sep 2013 18:22:52 +0100 |
aph |
Default to DeoptimizeWhenPatching.
|
| Thu, 12 Sep 2013 17:20:44 +0100 |
aph |
Fix the fix for operand_valid_for_float_immediate.
|
| Thu, 12 Sep 2013 16:04:07 +0100 |
aph |
Merge
|
| Thu, 12 Sep 2013 15:58:26 +0100 |
aph |
Fix operand_valid_for_float_immediate so that it returns false for negative zero.
|
| Thu, 12 Sep 2013 14:52:58 +0100 |
adinn |
Merge
|
| Wed, 11 Sep 2013 16:01:59 +0100 |
adinn |
make partialSubtypeCheck pass expected registers to check_klass_subtype_slow_path
|
| Thu, 12 Sep 2013 14:37:30 +0100 |
aph |
Merge
|
| Thu, 12 Sep 2013 14:30:24 +0100 |
aph |
Fix VerifyOops
|
| Thu, 12 Sep 2013 14:27:56 +0100 |
aph |
Use C fallback arraycopy for UseCompressedOops
|
| Wed, 11 Sep 2013 18:50:22 +0100 |
aph |
Fixes for compressed OOPs and VerifyOops
|
| Wed, 11 Sep 2013 13:32:05 +0100 |
aph |
Small optimization when unlocking
|
| Tue, 10 Sep 2013 16:37:19 +0100 |
aph |
merge
|
| Tue, 10 Sep 2013 16:37:00 +0100 |
aph |
Comment changes only
|
| Tue, 10 Sep 2013 18:06:43 +0100 |
adinn |
ensured that maxI and minI declare their use of the flags register
|
| Tue, 10 Sep 2013 18:05:09 +0100 |
adinn |
make cmovUI_reg_zero use correct flags register type
|
| Tue, 10 Sep 2013 11:28:04 +0100 |
adinn |
corrected logic for virtual call check
|
| Mon, 09 Sep 2013 16:52:46 +0100 |
adinn |
use correct count for number of int registers
|
| Mon, 09 Sep 2013 13:37:41 +0100 |
adinn |
relocated save and restore of cpu state from doCallout() to caller of run()
|
| Mon, 09 Sep 2013 13:16:30 +0100 |
adinn |
Merge
|
| Mon, 09 Sep 2013 13:06:31 +0100 |
adinn |
corrected declaration of FP regs 16-31 to make them save on call
|
| Fri, 06 Sep 2013 14:39:51 +0100 |
Edward Nevill edward nevill |
Fix ICache flusing problem found in RTSM
|
| Thu, 05 Sep 2013 18:16:19 +0100 |
adinn |
Merge
|
| Thu, 05 Sep 2013 18:15:53 +0100 |
adinn |
added missing CmpLTMask instruction
|
| Thu, 05 Sep 2013 18:14:43 +0100 |
adinn |
ensured call_VM_Leaf_base clears the pc it has just set
|
| Thu, 05 Sep 2013 18:12:47 +0100 |
adinn |
added csetw alias for csincw as per AArch64 spec definition
|
| Thu, 05 Sep 2013 18:10:35 +0100 |
adinn |
ensured reset_last_Java_frame calls clear the pc if they previously set it
|
| Wed, 04 Sep 2013 16:55:41 +0100 |
aph |
Temporary fix for single-stepping through method handle calls.
|
| Wed, 04 Sep 2013 15:41:31 +0100 |
aph |
merge
|
| Wed, 04 Sep 2013 15:40:38 +0100 |
aph |
Another fix for obj + long, this time for constant RHS.
|
| Wed, 04 Sep 2013 12:59:52 +0100 |
adinn |
Merge
|
| Wed, 04 Sep 2013 11:35:26 +0100 |
adinn |
made frame set/reset follow x86 and corrected some errors in frame code
|
| Wed, 04 Sep 2013 10:26:43 +0100 |
adinn |
corfrected 32/64 bit error in encoding of CompareAndSwapN
|
