| Mon, 10 Oct 2016 14:44:09 +0800 |
chenhaoxuan |
#4535 Loop unrolling for disjoint_short_copy
|
changeset |
files
|
| Sun, 09 Oct 2016 15:36:29 +0800 |
lifangyuan |
#4536 Added 128-bit memory access to generate_disjoint_short_copy.
|
changeset |
files
|
| Sat, 08 Oct 2016 09:31:28 -0400 |
fujie |
The parallel GC mark-phase should be MT-safe for 3A2000.
|
changeset |
files
|
| Thu, 29 Sep 2016 09:54:38 +0800 |
aoqi |
8049252: VerifyStack logic in Deoptimization::unpack_frames does not expect to see invoke bc at the top frame during normal deoptimization
|
changeset |
files
|
| Mon, 26 Sep 2016 16:01:19 -0400 |
fujie |
sync in oopDesc::set_mark for 3A2000 CPUs.
|
changeset |
files
|
| Mon, 26 Sep 2016 13:56:18 -0400 |
fujie |
Mark changes for 3A2000 only with the Use3A2000 flag.
|
changeset |
files
|
| Mon, 26 Sep 2016 13:04:07 -0400 |
fujie |
Add flag Use3A2000 in HotSpot.
|
changeset |
files
|
| Sun, 09 Oct 2016 14:48:55 +0800 |
aoqi |
merge
|
changeset |
files
|
| Tue, 20 Sep 2016 10:40:42 +0800 |
jiangshaofeng |
#4537 Rewrite generate_disjoint_int_oop_copy
|
changeset |
files
|
| Tue, 20 Sep 2016 11:48:21 +0800 |
jiangshaofeng |
#4537 Rewrite generate_disjoint_byte_copy
|
changeset |
files
|
| Sun, 25 Sep 2016 18:09:10 -0400 |
fujie |
Stability enhanced for 3A2000-4way.
|
changeset |
files
|
| Sun, 25 Sep 2016 17:38:33 -0400 |
fujie |
Sync after card marking in C2 compiler.
|
changeset |
files
|
| Mon, 26 Sep 2016 14:38:30 +0800 |
fujie |
Sync after card marking in interpreter.
|
changeset |
files
|
| Mon, 26 Sep 2016 14:21:21 +0800 |
fujie |
Sync in generate_conjoint_long_oop_copy and refine it.
|
changeset |
files
|
| Mon, 26 Sep 2016 14:01:22 +0800 |
fujie |
No sync when no oops are copied in generate_disjoint_int_oop_copy.
|
changeset |
files
|
| Mon, 26 Sep 2016 13:58:44 +0800 |
fujie |
Sync in generate_disjoint_long_oop_copy and refine it.
|
changeset |
files
|
| Mon, 26 Sep 2016 13:44:12 +0800 |
fujie |
Sync in generate_conjoint_int_oop_copy and refine it.
|
changeset |
files
|
| Mon, 26 Sep 2016 11:52:55 +0800 |
fujie |
Sync in generate_disjoint_int_oop_copy and refine it.
|
changeset |
files
|
| Thu, 22 Sep 2016 07:41:28 -0400 |
fujie |
Fix a bug while copying oops in the heap with -XX:-ReduceInitialCardMarks.
|
changeset |
files
|
| Sun, 18 Sep 2016 13:43:10 +0800 |
Jin |
os::active_processor_count() is modified to return the number of processors numactl-binded.
|
changeset |
files
|
| Sat, 17 Sep 2016 03:44:08 -0400 |
fujie |
Sync before write on 3A2000 (Follow changeset 9f90c85c6815).
|
changeset |
files
|
| Mon, 12 Sep 2016 13:59:25 -0400 |
fujie |
Fix a bug when decoding the compressed klass oops.
|
changeset |
files
|
| Mon, 12 Sep 2016 13:50:08 -0400 |
fujie |
Remove redundant syncs.
|
changeset |
files
|
| Mon, 12 Sep 2016 13:41:43 -0400 |
fujie |
Sync before updating the card marks (must be MT-safe on 3A2000).
|
changeset |
files
|
| Mon, 12 Sep 2016 13:26:09 -0400 |
fujie |
Add debugging infomation for C2 compiler.
|
changeset |
files
|
| Thu, 08 Sep 2016 09:59:34 +0800 |
aoqi |
Fixed an n64 ABI incompatibility in print_reg.
|
changeset |
files
|
| Wed, 24 Aug 2016 15:38:31 +0800 |
fujie |
Add salI_RegL2I_imm in mips_64.ad
|
changeset |
files
|
| Tue, 23 Aug 2016 17:00:34 +0800 |
fujie |
Refine the implementation of cmpLTMask_immI0.
|
changeset |
files
|
| Tue, 23 Aug 2016 15:23:48 +0800 |
fujie |
Add slrL_Reg_immI_0_31 and slrL_Reg_immI_32_63 in mips_64.ad
|
changeset |
files
|
| Tue, 23 Aug 2016 14:22:12 +0800 |
fujie |
Add storeI_convL2I in mips_64.add. crypto.rsa is 13% up.
|
changeset |
files
|
| Tue, 23 Aug 2016 10:45:23 +0800 |
fujie |
Add immL16 im mips_64.ad
|
changeset |
files
|
| Tue, 23 Aug 2016 09:21:30 +0800 |
fujie |
Add loadUS_convI2L in mips_64.ad
|
changeset |
files
|
| Tue, 23 Aug 2016 09:16:04 +0800 |
fujie |
Add loadS_convI2L in mips_64.ad
|
changeset |
files
|
| Tue, 23 Aug 2016 09:12:34 +0800 |
fujie |
Add loadUB_convI2L in mips_64.ad
|
changeset |
files
|
| Tue, 23 Aug 2016 09:07:41 +0800 |
fujie |
Add loadB_convI2L in mips_64.ad
|
changeset |
files
|
| Mon, 22 Aug 2016 17:51:21 +0800 |
fujie |
Add cmpLTMask_immI0.
|
changeset |
files
|
| Mon, 22 Aug 2016 17:29:54 +0800 |
fujie |
Add andL_Reg_imm_0_65535 for AndL.
|
changeset |
files
|
| Thu, 25 Aug 2016 22:31:58 +0800 |
jiangshaofeng |
#4428: Use gsdmult to optimize the lmul template of the interpreter.
|
changeset |
files
|
| Fri, 26 Aug 2016 20:31:57 +0800 |
lifangyuan |
#4460:Performance of long integer mod operation for intepreter is 11.7% up.
|
changeset |
files
|
| Thu, 25 Aug 2016 23:57:31 +0800 |
chenhaoxuan |
#4459: Performance of long integer division for intepreter is 13% up.
|
changeset |
files
|
| Thu, 25 Aug 2016 17:34:45 +0800 |
fujie |
For safety reason, disable gslbx for Loongson processors.
|
changeset |
files
|
| Thu, 25 Aug 2016 14:46:18 +0800 |
aoqi |
#4476 #4461: Fix a bug related to float comparison.
|
changeset |
files
|
| Thu, 18 Aug 2016 15:57:28 +0800 |
fujie |
Remove sync in PrefetchAllocation.
|
changeset |
files
|
| Thu, 18 Aug 2016 15:23:16 +0800 |
fujie |
Remove unnecessary sync instructions in mips_64.ad
|
changeset |
files
|
| Thu, 18 Aug 2016 14:17:07 +0800 |
fujie |
Enable gsswx in store_I_reg_enc for Loongson processors.
|
changeset |
files
|
| Thu, 18 Aug 2016 13:51:09 +0800 |
fujie |
Enable gssdxc1 in store_D_reg_enc for Loongson processors.
|
changeset |
files
|
| Thu, 18 Aug 2016 13:38:41 +0800 |
fujie |
Enable gsldxc1 in load_D_enc for Loongson processors.
|
changeset |
files
|
| Thu, 18 Aug 2016 11:54:02 +0800 |
fujie |
Fix a bug in the definitions of (gslwxc1/gsldxc1/gsswxc1/gssdxc1): Register --> FloatRegister.
|
changeset |
files
|
| Thu, 18 Aug 2016 11:50:18 +0800 |
fujie |
Enable gsswxc1 for Loongson Processors.
|
changeset |
files
|
| Thu, 18 Aug 2016 11:41:31 +0800 |
fujie |
Enable gslwxc1 in load_F_enc for Loongson processors.
|
changeset |
files
|
| Thu, 18 Aug 2016 11:33:16 +0800 |
fujie |
Add gsswxc1 and gssdxc1 instructions.
|
changeset |
files
|
| Thu, 18 Aug 2016 11:29:03 +0800 |
fujie |
Add gslwxc1 and gsldxc1 instruction.
|
changeset |
files
|
| Thu, 18 Aug 2016 11:16:54 +0800 |
fujie |
Add addL_RegI2L_Reg and addL_Reg_RegI2L in mips_64.ad
|
changeset |
files
|
| Thu, 18 Aug 2016 11:12:52 +0800 |
fujie |
Add subL_Reg_RegI2L and subL_RegI2L_Reg in mips_64.ad
|
changeset |
files
|
| Thu, 18 Aug 2016 10:16:47 +0800 |
fujie |
Add operand baseIndexOffset8_convI2L in mips_64.add.
|
changeset |
files
|
| Thu, 18 Aug 2016 09:58:51 +0800 |
fujie |
Enable gsshx in store_C_reg_enc for Loongson processors.
|
changeset |
files
|
| Thu, 18 Aug 2016 09:46:58 +0800 |
fujie |
Enable gslbx in load_B_enc for Loongson processors.
|
changeset |
files
|
| Wed, 17 Aug 2016 17:20:44 +0800 |
fujie |
Performance of integer array operation is about 8% up.
|
changeset |
files
|
| Wed, 17 Aug 2016 15:26:28 +0800 |
fujie |
Matching for baseIndexOffset8 is OK.
|
changeset |
files
|
| Tue, 16 Aug 2016 15:31:40 +0800 |
fujie |
Add gslhx/gslbx instructions for Loongson processors.
|
changeset |
files
|
