December 2017 - aarch64-port-dev

RFR: 8191954: AArch64: disable UseCISCSpill in C2
by Ningsheng Jian 20 Mar '18

20 Mar '18

Hi, Please help to review. JBS: https://bugs.openjdk.java.net/browse/JDK-8191954 Webrev: http://cr.openjdk.java.net/~njian/8191954/webrev.00/ This sets UseCISCSpill to false on AArch64, as we actually don't use that. jtreg tests passed with fastdebug build. Thanks, Ningsheng

4 7

[RFC] ldp/stp peephole optimizations
by Zhongwei Yao 30 Jan '18

30 Jan '18

Hi, We are planning to add AArch64 LDP/STP (load/store pair of registers) support in C2 code-gen for better performance. I think the LDP/STP can be used in following cases: A). For register spill/unspill. We've observed many sequential single stack load/store patterns in SPECjbb C2 generated code. B). Besides spilling, LDP is also not generated generally for multiple LoadI/LoadL nodes. Is there any risk (e.g. implicit check?) for combing them together, apart from alignment issue? I think peephole is the best fit for above optimization (gcc/llvm also has such peephole optimization). However, current peephole rules in C2 compiler is very limited and I doubt whether it really takes effect - AArch64 has disabled peephole optimizations. x86 has enabled it, but the instruction sequences to be matched by the rules seems to be very uncommon. To address issue A), since current spill/unspill are handled by common MachSpillCopyNode, I was thinking if we could add peephole rule to match MachSpillCopyNode, but MachSpillCopyNode has no operands (e.g. mem, src, dst) like ordinary instruct defined in aarch64.ad. Even we may extract them (mem, src, dst) like in MachSpillCopyNode::implementation(), and even we can extend current peephole rule grammar, expressing such extraction in peephole's grammar is complex. So I prefer adding following manually defined method peephole() to MachSpillCopyNode: virtual MachNode *peephole(Block *block, int block_index, PhaseRegAlloc *ra_, int &deleted); This makes the patch relative simple. My prototype patch for A) (still some TODOs and hardcodes, but it works fine): http://cr.openjdk.java.net/~zyao/RFC_A/ To address issue B) is somewhat complicated, we need to extend current peephole rule syntax, as I don't think current simple syntax works for any useful peephole optimizations like ldp/stp opt. My extended syntax - at least works for ldp/stp optimizations: ------ peepmatch ( loadI loadI ); peepconstraint (0.mem$base == 1.mem$base, 0.mem$scale == 1.mem$scale, 0.mem$disp - 4 == 1.mem$disp, 0.dst != 1.dst); // new grammar is described below. peepreplace (loadPairI(1.mem 1.mem)) ------ But for loadPairI, it is hard to express in current instruct semantic. Because current instruct in aarch64.ad is defined by a match rule. The match rule is an expression tree and made of Ideal Node. However, LDP instruction doesn't have Ideal Node (say LoadPair) to match. And adding load pair node to arch-independent Ideal node seems strange. My proposed solution is: add a special arch dependent operand like iRegIpair: ------ operand iRegIpair(iRegI reg1, iRegI reg2) %{ constraint(ALLOC_IN_RC(any_reg32)); op_cost(0); format %{ "pair: reg1, reg2"%}; // hard coded format for now. interface(REG_INTER); %} ------ This needs to update ADLC to support iRegIpair operand. Because unlike current operand which has 1 register, iRegIpair has 2. Then use it as loadPairI's operand type like: ------ instruct loadPairI(indOffI mem, iRegIpair dst) %{ match(Set dst mem); //no Ideal Node in match rule. ... %} ------ Then we can use loadPairI in peephole rule's "peepreplace". Since only constraints between operands are supported in peephole rule. But to check whether the adjacent loads are loaded from adjacent memory address, we need to check operand's member, like (0.mem$disp - 4 == 1.mem$disp), My solution is: add new grammar like 0.mem$disp to extract member in operand in ADLC (peep_constraint_parse()). Another issue for peephole optimization is that it only matches adjacent instructions in the same basic block. This leads to many missing matches when loads are not scheduled to adjacent. So I propose to delay peephole phase to the place just before final code emit (the fill_buffer() function). This place is after instruction scheduling. So after instruction scheduling, we could match more adjacent loads. My draft patch to address B) is at: http://cr.openjdk.java.net/~zyao/RFC_B/ What do you think? Welcome any feedback! -- Best regards, Zhongwei

4 9

RFR(S): 8194256 - AARCH64: SIMD shift instructions are incorrectly encoded
by Dmitrij Pochepko 18 Jan '18

18 Jan '18

Hi all, please review small patch for 8194256 - AARCH64: SIMD shift instructions are incorrectly encoded I've noticed SIMD shift instructions wrong encoding when trying to use it. An intrinsic I was working on, generated incorrect assembly code with wrong shift value. Existing code just copy "shift" bits into instruction bits(immh:immb), however, according to spec, it should be encoded as follows: SIMD type <T>: 8B when immh = 0001 , Q = 0 16B when immh = 0001 , Q = 1 4H when immh = 001x , Q = 0 8H when immh = 001x , Q = 1 2S when immh = 01xx , Q = 0 4S when immh = 01xx , Q = 1 2D when immh = 1xxx , Q = 1 <shift> is encoded as follows: for ushr and sshr: (16-UInt(immh:immb)) when immh = 0001 (32-UInt(immh:immb)) when immh = 001x (64-UInt(immh:immb)) when immh = 01xx (128-UInt(immh:immb)) when immh = 1xxx for shl: (UInt(immh:immb)-8) when immh = 0001 (UInt(immh:immb)-16) when immh = 001x (UInt(immh:immb)-32) when immh = 01xx (UInt(immh:immb)-64) when immh = 1xxx So, I've modified respective instruction generation code. webrev: http://cr.openjdk.java.net/~dpochepk/8194256/webrev.01/ CR: https://bugs.openjdk.java.net/browse/JDK-8194256 I've checked this patch by generating intrinsic which use these instructions and verified assembly code. Thanks, Dmitrij

3 4

RFR: 8193266: AArch64: TestOptionsWithRanges.java SIGSEGV
by Stuart Monteith 04 Jan '18

04 Jan '18

Hello, The Aarch64 backend decompresses klass pointers using XOR if the size of the compressed classes space is less then the base address. However, in JDK 10 CompressedClassSpaceSize doesn't represent the range of the compressed shared classes area, which is set to 4GB. https://bugs.openjdk.java.net/browse/JDK-8193266 http://cr.openjdk.java.net/~smonteith/8193266/webrev/ I've speculatively put up a patch which deals with the issue on Aarch64. While I understand the problem with the generated aarch64 code, I don't believe understand the CDS code sufficiently at this point to determine if there is a bug within the CDS code regarding the CompressedClassSpaceSize parameter. Thanks, Stuart

3 12

JTREG, JCStress, SPECjbb2015 and Hadoop/Terasort results for OpenJDK JDK on AArch64
by ci_notify＠linaro.org 28 Dec '17

28 Dec '17

This is a summary of the JTREG test results =========================================== The build and test results are cycled every 15 days. For detailed information on the test output please refer to: http://openjdk.linaro.org/jdkX/openjdk-jtreg-nightly-tests/summary/2017/361… ------------------------------------------------------------------------------- client-release/hotspot ------------------------------------------------------------------------------- Build 0: aarch64/2017/nov/25 pass: 1,456; fail: 12; error: 1 Build 1: aarch64/2017/nov/27 pass: 1,457; fail: 11; error: 1 Build 2: aarch64/2017/nov/29 pass: 1,546; fail: 14 Build 3: aarch64/2017/dec/01 pass: 1,575; fail: 11 Build 4: aarch64/2017/dec/03 pass: 1,580; fail: 11 Build 5: aarch64/2017/dec/05 pass: 1,580; fail: 11 Build 6: aarch64/2017/dec/07 pass: 1,578; fail: 12; error: 1 Build 7: aarch64/2017/dec/09 pass: 1,584; fail: 10 Build 8: aarch64/2017/dec/11 pass: 1,584; fail: 10 Build 9: aarch64/2017/dec/13 pass: 1,585; fail: 9 Build 10: aarch64/2017/dec/15 pass: 1,588; fail: 10 Build 11: aarch64/2017/dec/17 pass: 1,597; fail: 11; error: 1 Build 12: aarch64/2017/dec/19 pass: 1,599; fail: 11 Build 13: aarch64/2017/dec/23 pass: 1,599; fail: 12 Build 14: aarch64/2017/dec/27 pass: 1,598; fail: 13 1 fatal errors were detected; please follow the link above for more detail. ------------------------------------------------------------------------------- client-release/jdk ------------------------------------------------------------------------------- Build 0: aarch64/2017/nov/25 pass: 7,531; fail: 697; error: 27 Build 1: aarch64/2017/nov/27 pass: 7,507; fail: 730; error: 18 Build 2: aarch64/2017/nov/29 pass: 7,501; fail: 737; error: 18 Build 3: aarch64/2017/dec/01 pass: 7,501; fail: 735; error: 20 Build 4: aarch64/2017/dec/03 pass: 7,532; fail: 710; error: 23 Build 5: aarch64/2017/dec/05 pass: 7,510; fail: 732; error: 23 Build 6: aarch64/2017/dec/07 pass: 7,527; fail: 715; error: 23 Build 7: aarch64/2017/dec/09 pass: 7,558; fail: 727; error: 22 Build 8: aarch64/2017/dec/11 pass: 7,559; fail: 732; error: 19 Build 9: aarch64/2017/dec/13 pass: 7,558; fail: 725; error: 27 Build 10: aarch64/2017/dec/15 pass: 7,569; fail: 741; error: 25 Build 11: aarch64/2017/dec/17 pass: 7,591; fail: 719; error: 26 Build 12: aarch64/2017/dec/19 pass: 7,572; fail: 740; error: 24 Build 13: aarch64/2017/dec/23 pass: 7,581; fail: 734; error: 20 Build 14: aarch64/2017/dec/27 pass: 7,575; fail: 736; error: 24 ------------------------------------------------------------------------------- client-release/langtools ------------------------------------------------------------------------------- Build 0: aarch64/2017/nov/25 pass: 3,818; fail: 4; error: 1 Build 1: aarch64/2017/nov/27 pass: 3,814; fail: 5; error: 4 Build 2: aarch64/2017/nov/29 pass: 3,820; fail: 3 Build 3: aarch64/2017/dec/01 pass: 3,818; fail: 3; error: 2 Build 4: aarch64/2017/dec/03 pass: 3,829; fail: 4; error: 1 Build 5: aarch64/2017/dec/05 pass: 3,828; fail: 4; error: 2 Build 6: aarch64/2017/dec/07 pass: 3,828; fail: 3; error: 3 Build 7: aarch64/2017/dec/09 pass: 3,830; fail: 4; error: 2 Build 8: aarch64/2017/dec/11 pass: 3,830; fail: 4; error: 1 Build 9: aarch64/2017/dec/13 pass: 3,831; fail: 4 Build 10: aarch64/2017/dec/15 pass: 3,835; fail: 3; error: 3 Build 11: aarch64/2017/dec/17 pass: 3,820; fail: 3; error: 3 Build 12: aarch64/2017/dec/19 pass: 3,819; fail: 3; error: 4 Build 13: aarch64/2017/dec/23 pass: 3,818; fail: 5; error: 4 Build 14: aarch64/2017/dec/27 pass: 3,821; fail: 5; error: 1 ------------------------------------------------------------------------------- server-release/hotspot ------------------------------------------------------------------------------- Build 0: aarch64/2017/nov/25 pass: 1,460; fail: 13; error: 1 Build 1: aarch64/2017/nov/27 pass: 1,460; fail: 13; error: 1 Build 2: aarch64/2017/nov/29 pass: 1,550; fail: 15 Build 3: aarch64/2017/dec/01 pass: 1,580; fail: 11 Build 4: aarch64/2017/dec/03 pass: 1,584; fail: 11; error: 1 Build 5: aarch64/2017/dec/05 pass: 1,585; fail: 10; error: 1 Build 6: aarch64/2017/dec/07 pass: 1,584; fail: 11; error: 1 Build 7: aarch64/2017/dec/09 pass: 1,588; fail: 10; error: 1 Build 8: aarch64/2017/dec/11 pass: 1,589; fail: 9; error: 1 Build 9: aarch64/2017/dec/13 pass: 1,589; fail: 9; error: 1 Build 10: aarch64/2017/dec/15 pass: 1,592; fail: 11 Build 11: aarch64/2017/dec/17 pass: 1,604; fail: 10 Build 12: aarch64/2017/dec/19 pass: 1,605; fail: 10 Build 13: aarch64/2017/dec/23 pass: 1,604; fail: 11; error: 1 Build 14: aarch64/2017/dec/27 pass: 1,603; fail: 12; error: 1 1 fatal errors were detected; please follow the link above for more detail. ------------------------------------------------------------------------------- server-release/jdk ------------------------------------------------------------------------------- Build 0: aarch64/2017/nov/25 pass: 7,511; fail: 715; error: 29 Build 1: aarch64/2017/nov/27 pass: 7,528; fail: 705; error: 22 Build 2: aarch64/2017/nov/29 pass: 7,512; fail: 719; error: 25 Build 3: aarch64/2017/dec/01 pass: 7,526; fail: 708; error: 22 Build 4: aarch64/2017/dec/03 pass: 7,530; fail: 710; error: 25 Build 5: aarch64/2017/dec/05 pass: 7,529; fail: 716; error: 20 Build 6: aarch64/2017/dec/07 pass: 7,543; fail: 690; error: 32 Build 7: aarch64/2017/dec/09 pass: 7,570; fail: 716; error: 21 Build 8: aarch64/2017/dec/11 pass: 7,584; fail: 705; error: 21 Build 9: aarch64/2017/dec/13 pass: 7,577; fail: 713; error: 20 Build 10: aarch64/2017/dec/15 pass: 7,600; fail: 711; error: 24 Build 11: aarch64/2017/dec/17 pass: 7,593; fail: 721; error: 22 Build 12: aarch64/2017/dec/19 pass: 7,586; fail: 726; error: 24 Build 13: aarch64/2017/dec/23 pass: 7,596; fail: 717; error: 22 Build 14: aarch64/2017/dec/27 pass: 7,589; fail: 721; error: 25 ------------------------------------------------------------------------------- server-release/langtools ------------------------------------------------------------------------------- Build 0: aarch64/2017/nov/25 pass: 3,819; fail: 4 Build 1: aarch64/2017/nov/27 pass: 3,819; fail: 3; error: 1 Build 2: aarch64/2017/nov/29 pass: 3,819; fail: 3; error: 1 Build 3: aarch64/2017/dec/01 pass: 3,819; fail: 4 Build 4: aarch64/2017/dec/03 pass: 3,829; fail: 3; error: 2 Build 5: aarch64/2017/dec/05 pass: 3,829; fail: 4; error: 1 Build 6: aarch64/2017/dec/07 pass: 3,829; fail: 3; error: 2 Build 7: aarch64/2017/dec/09 pass: 3,832; fail: 3; error: 1 Build 8: aarch64/2017/dec/11 pass: 3,830; fail: 3; error: 2 Build 9: aarch64/2017/dec/13 pass: 3,830; fail: 4; error: 1 Build 10: aarch64/2017/dec/15 pass: 3,833; fail: 3; error: 5 Build 11: aarch64/2017/dec/17 pass: 3,822; fail: 3; error: 1 Build 12: aarch64/2017/dec/19 pass: 3,817; fail: 3; error: 6 Build 13: aarch64/2017/dec/23 pass: 3,819; fail: 5; error: 3 Build 14: aarch64/2017/dec/27 pass: 3,820; fail: 5; error: 2 Previous results can be found here: http://openjdk.linaro.org/jdkX/openjdk-jtreg-nightly-tests/index.html SPECjbb2015 composite regression test completed =============================================== This test measures the relative performance of the server compiler running the SPECjbb2015 composite tests and compares the performance against the baseline performance of the server compiler taken on 2016-11-21. In accordance with [1], the SPECjbb2015 tests are run on a system which is not production ready and does not meet all the requirements for publishing compliant results. The numbers below shall be treated as non-compliant (nc) and are for experimental purposes only. Relative performance: Server max-jOPS (nc): 1.01x Relative performance: Server critical-jOPS (nc): 0.86x Details of the test setup and historical results may be found here: http://openjdk.linaro.org/jdkX/SPECjbb2015-results/ [1] http://www.spec.org/fairuse.html#Academic Regression test Hadoop-Terasort completed ========================================= This test measures the performance of the server and client compilers running Hadoop sorting a 1GB file using Terasort and compares the performance against the baseline performance of the Zero interpreter and against the baseline performance of the client and server compilers on 2014-04-01. Relative performance: Zero: 1.0, Client: 70.93, Server: 114.76 Client 70.93 / Client 2014-04-01 (43.00): 1.65x Server 114.76 / Server 2014-04-01 (71.00): 1.62x Details of the test setup and historical results may be found here: http://openjdk.linaro.org/jdkX/hadoop-terasort-benchmark-results/ This is a summary of the jcstress test results ============================================== The build and test results are cycled every 15 days. 2017-11-26 pass rate: 11558/11560, results: http://openjdk.linaro.org/jdkX/jcstress-nightly-runs/2017/329/results/ 2017-11-28 pass rate: 11559/11560, results: http://openjdk.linaro.org/jdkX/jcstress-nightly-runs/2017/331/results/ 2017-11-30 pass rate: 11557/11560, results: http://openjdk.linaro.org/jdkX/jcstress-nightly-runs/2017/333/results/ 2017-12-02 pass rate: 11557/11560, results: http://openjdk.linaro.org/jdkX/jcstress-nightly-runs/2017/335/results/ 2017-12-04 pass rate: 11557/11560, results: http://openjdk.linaro.org/jdkX/jcstress-nightly-runs/2017/337/results/ 2017-12-06 pass rate: 11558/11560, results: http://openjdk.linaro.org/jdkX/jcstress-nightly-runs/2017/339/results/ 2017-12-08 pass rate: 11556/11560, results: http://openjdk.linaro.org/jdkX/jcstress-nightly-runs/2017/341/results/ 2017-12-10 pass rate: 11559/11560, results: http://openjdk.linaro.org/jdkX/jcstress-nightly-runs/2017/343/results/ 2017-12-12 pass rate: 11558/11560, results: http://openjdk.linaro.org/jdkX/jcstress-nightly-runs/2017/345/results/ 2017-12-14 pass rate: 11558/11560, results: http://openjdk.linaro.org/jdkX/jcstress-nightly-runs/2017/347/results/ 2017-12-16 pass rate: 11557/11560, results: http://openjdk.linaro.org/jdkX/jcstress-nightly-runs/2017/349/results/ 2017-12-18 pass rate: 11558/11560, results: http://openjdk.linaro.org/jdkX/jcstress-nightly-runs/2017/351/results/ 2017-12-20 pass rate: 11559/11560, results: http://openjdk.linaro.org/jdkX/jcstress-nightly-runs/2017/353/results/ 2017-12-24 pass rate: 11557/11560, results: http://openjdk.linaro.org/jdkX/jcstress-nightly-runs/2017/357/results/ 2017-12-28 pass rate: 11558/11560, results: http://openjdk.linaro.org/jdkX/jcstress-nightly-runs/2017/361/results/ For detailed information on the test output please refer to: http://openjdk.linaro.org/jdkX/jcstress-nightly-runs/

1 0

Build failure: invalid register name for ‘r_exchange_value’
by Aleksey Shipilev 24 Dec '17

24 Dec '17

Hi, Seeing weird build failures in aarch64/jdk8u and aarch64/jdk8u-shenandoah during aarch64 cross-compilation: /pool/buildbot/slaves/sobornost/jdk8u-rh-int/build/hotspot/src/os_cpu/linux_aarch64/vm/atomic_linux_aarch64.inline.hpp: In member function ‘void AccessFlags::atomic_set_bits(jint)’: /pool/buildbot/slaves/sobornost/jdk8u-rh-int/build/hotspot/src/os_cpu/linux_aarch64/vm/atomic_linux_aarch64.inline.hpp:128:18: error: invalid register name for ‘r_exchange_value’ register jint r_exchange_value asm("w0") = exchange_value; ^~~~~~~~~~~~~~~~ Seems to be coming from: changeset: 9155:7fc149d14601 user: enevill date: Tue Mar 29 10:07:54 2016 +0000 summary: 8151775: aarch64: add support for 8.1 LSE atomic operations Bugreport mentions the fix had come to jdk9: http://hg.openjdk.java.net/jdk9/jdk9/hotspot/rev/03fb00b96355 What is concerning is that the change in aarch64/8u is *different*, it has the assembly block where the build fails: http://hg.openjdk.java.net/aarch64-port/jdk8u/hotspot/rev/7fc149d14601 What's going on here? Shouldn't aarch64-port version be closer to jdk9? Edward? Thanks, -Aleksey

2 3

JTREG, JCStress, SPECjbb2015 and Hadoop/Terasort results for OpenJDK JDK on AArch64
by ci_notify＠linaro.org 24 Dec '17

24 Dec '17

This is a summary of the JTREG test results =========================================== The build and test results are cycled every 15 days. For detailed information on the test output please refer to: http://openjdk.linaro.org/jdkX/openjdk-jtreg-nightly-tests/summary/2017/357… ------------------------------------------------------------------------------- client-release/hotspot ------------------------------------------------------------------------------- Build 0: aarch64/2017/nov/23 pass: 1,448; fail: 12 Build 1: aarch64/2017/nov/25 pass: 1,456; fail: 12; error: 1 Build 2: aarch64/2017/nov/27 pass: 1,457; fail: 11; error: 1 Build 3: aarch64/2017/nov/29 pass: 1,546; fail: 14 Build 4: aarch64/2017/dec/01 pass: 1,575; fail: 11 Build 5: aarch64/2017/dec/03 pass: 1,580; fail: 11 Build 6: aarch64/2017/dec/05 pass: 1,580; fail: 11 Build 7: aarch64/2017/dec/07 pass: 1,578; fail: 12; error: 1 Build 8: aarch64/2017/dec/09 pass: 1,584; fail: 10 Build 9: aarch64/2017/dec/11 pass: 1,584; fail: 10 Build 10: aarch64/2017/dec/13 pass: 1,585; fail: 9 Build 11: aarch64/2017/dec/15 pass: 1,588; fail: 10 Build 12: aarch64/2017/dec/17 pass: 1,597; fail: 11; error: 1 Build 13: aarch64/2017/dec/19 pass: 1,599; fail: 11 Build 14: aarch64/2017/dec/23 pass: 1,599; fail: 12 1 fatal errors were detected; please follow the link above for more detail. ------------------------------------------------------------------------------- client-release/jdk ------------------------------------------------------------------------------- Build 0: aarch64/2017/nov/23 pass: 7,505; fail: 728; error: 22 Build 1: aarch64/2017/nov/25 pass: 7,531; fail: 697; error: 27 Build 2: aarch64/2017/nov/27 pass: 7,507; fail: 730; error: 18 Build 3: aarch64/2017/nov/29 pass: 7,501; fail: 737; error: 18 Build 4: aarch64/2017/dec/01 pass: 7,501; fail: 735; error: 20 Build 5: aarch64/2017/dec/03 pass: 7,532; fail: 710; error: 23 Build 6: aarch64/2017/dec/05 pass: 7,510; fail: 732; error: 23 Build 7: aarch64/2017/dec/07 pass: 7,527; fail: 715; error: 23 Build 8: aarch64/2017/dec/09 pass: 7,558; fail: 727; error: 22 Build 9: aarch64/2017/dec/11 pass: 7,559; fail: 732; error: 19 Build 10: aarch64/2017/dec/13 pass: 7,558; fail: 725; error: 27 Build 11: aarch64/2017/dec/15 pass: 7,569; fail: 741; error: 25 Build 12: aarch64/2017/dec/17 pass: 7,591; fail: 719; error: 26 Build 13: aarch64/2017/dec/19 pass: 7,572; fail: 740; error: 24 Build 14: aarch64/2017/dec/23 pass: 7,581; fail: 734; error: 20 1 fatal errors were detected; please follow the link above for more detail. ------------------------------------------------------------------------------- client-release/langtools ------------------------------------------------------------------------------- Build 0: aarch64/2017/nov/23 pass: 3,817; fail: 5; error: 1 Build 1: aarch64/2017/nov/25 pass: 3,818; fail: 4; error: 1 Build 2: aarch64/2017/nov/27 pass: 3,814; fail: 5; error: 4 Build 3: aarch64/2017/nov/29 pass: 3,820; fail: 3 Build 4: aarch64/2017/dec/01 pass: 3,818; fail: 3; error: 2 Build 5: aarch64/2017/dec/03 pass: 3,829; fail: 4; error: 1 Build 6: aarch64/2017/dec/05 pass: 3,828; fail: 4; error: 2 Build 7: aarch64/2017/dec/07 pass: 3,828; fail: 3; error: 3 Build 8: aarch64/2017/dec/09 pass: 3,830; fail: 4; error: 2 Build 9: aarch64/2017/dec/11 pass: 3,830; fail: 4; error: 1 Build 10: aarch64/2017/dec/13 pass: 3,831; fail: 4 Build 11: aarch64/2017/dec/15 pass: 3,835; fail: 3; error: 3 Build 12: aarch64/2017/dec/17 pass: 3,820; fail: 3; error: 3 Build 13: aarch64/2017/dec/19 pass: 3,819; fail: 3; error: 4 Build 14: aarch64/2017/dec/23 pass: 3,818; fail: 5; error: 4 ------------------------------------------------------------------------------- server-release/hotspot ------------------------------------------------------------------------------- Build 0: aarch64/2017/nov/23 pass: 1,452; fail: 13 Build 1: aarch64/2017/nov/25 pass: 1,460; fail: 13; error: 1 Build 2: aarch64/2017/nov/27 pass: 1,460; fail: 13; error: 1 Build 3: aarch64/2017/nov/29 pass: 1,550; fail: 15 Build 4: aarch64/2017/dec/01 pass: 1,580; fail: 11 Build 5: aarch64/2017/dec/03 pass: 1,584; fail: 11; error: 1 Build 6: aarch64/2017/dec/05 pass: 1,585; fail: 10; error: 1 Build 7: aarch64/2017/dec/07 pass: 1,584; fail: 11; error: 1 Build 8: aarch64/2017/dec/09 pass: 1,588; fail: 10; error: 1 Build 9: aarch64/2017/dec/11 pass: 1,589; fail: 9; error: 1 Build 10: aarch64/2017/dec/13 pass: 1,589; fail: 9; error: 1 Build 11: aarch64/2017/dec/15 pass: 1,592; fail: 11 Build 12: aarch64/2017/dec/17 pass: 1,604; fail: 10 Build 13: aarch64/2017/dec/19 pass: 1,605; fail: 10 Build 14: aarch64/2017/dec/23 pass: 1,604; fail: 11; error: 1 1 fatal errors were detected; please follow the link above for more detail. ------------------------------------------------------------------------------- server-release/jdk ------------------------------------------------------------------------------- Build 0: aarch64/2017/nov/23 pass: 7,502; fail: 728; error: 25 Build 1: aarch64/2017/nov/25 pass: 7,511; fail: 715; error: 29 Build 2: aarch64/2017/nov/27 pass: 7,528; fail: 705; error: 22 Build 3: aarch64/2017/nov/29 pass: 7,512; fail: 719; error: 25 Build 4: aarch64/2017/dec/01 pass: 7,526; fail: 708; error: 22 Build 5: aarch64/2017/dec/03 pass: 7,530; fail: 710; error: 25 Build 6: aarch64/2017/dec/05 pass: 7,529; fail: 716; error: 20 Build 7: aarch64/2017/dec/07 pass: 7,543; fail: 690; error: 32 Build 8: aarch64/2017/dec/09 pass: 7,570; fail: 716; error: 21 Build 9: aarch64/2017/dec/11 pass: 7,584; fail: 705; error: 21 Build 10: aarch64/2017/dec/13 pass: 7,577; fail: 713; error: 20 Build 11: aarch64/2017/dec/15 pass: 7,600; fail: 711; error: 24 Build 12: aarch64/2017/dec/17 pass: 7,593; fail: 721; error: 22 Build 13: aarch64/2017/dec/19 pass: 7,586; fail: 726; error: 24 Build 14: aarch64/2017/dec/23 pass: 7,596; fail: 717; error: 22 ------------------------------------------------------------------------------- server-release/langtools ------------------------------------------------------------------------------- Build 0: aarch64/2017/nov/23 pass: 3,816; fail: 4; error: 3 Build 1: aarch64/2017/nov/25 pass: 3,819; fail: 4 Build 2: aarch64/2017/nov/27 pass: 3,819; fail: 3; error: 1 Build 3: aarch64/2017/nov/29 pass: 3,819; fail: 3; error: 1 Build 4: aarch64/2017/dec/01 pass: 3,819; fail: 4 Build 5: aarch64/2017/dec/03 pass: 3,829; fail: 3; error: 2 Build 6: aarch64/2017/dec/05 pass: 3,829; fail: 4; error: 1 Build 7: aarch64/2017/dec/07 pass: 3,829; fail: 3; error: 2 Build 8: aarch64/2017/dec/09 pass: 3,832; fail: 3; error: 1 Build 9: aarch64/2017/dec/11 pass: 3,830; fail: 3; error: 2 Build 10: aarch64/2017/dec/13 pass: 3,830; fail: 4; error: 1 Build 11: aarch64/2017/dec/15 pass: 3,833; fail: 3; error: 5 Build 12: aarch64/2017/dec/17 pass: 3,822; fail: 3; error: 1 Build 13: aarch64/2017/dec/19 pass: 3,817; fail: 3; error: 6 Build 14: aarch64/2017/dec/23 pass: 3,819; fail: 5; error: 3 Previous results can be found here: http://openjdk.linaro.org/jdkX/openjdk-jtreg-nightly-tests/index.html SPECjbb2015 composite regression test completed =============================================== This test measures the relative performance of the server compiler running the SPECjbb2015 composite tests and compares the performance against the baseline performance of the server compiler taken on 2016-11-21. In accordance with [1], the SPECjbb2015 tests are run on a system which is not production ready and does not meet all the requirements for publishing compliant results. The numbers below shall be treated as non-compliant (nc) and are for experimental purposes only. Relative performance: Server max-jOPS (nc): 1.00x Relative performance: Server critical-jOPS (nc): 0.85x Details of the test setup and historical results may be found here: http://openjdk.linaro.org/jdkX/SPECjbb2015-results/ [1] http://www.spec.org/fairuse.html#Academic Regression test Hadoop-Terasort completed ========================================= This test measures the performance of the server and client compilers running Hadoop sorting a 1GB file using Terasort and compares the performance against the baseline performance of the Zero interpreter and against the baseline performance of the client and server compilers on 2014-04-01. Relative performance: Zero: 1.0, Client: 69.88, Server: 111.14 Client 69.88 / Client 2014-04-01 (43.00): 1.63x Server 111.14 / Server 2014-04-01 (71.00): 1.57x Details of the test setup and historical results may be found here: http://openjdk.linaro.org/jdkX/hadoop-terasort-benchmark-results/ This is a summary of the jcstress test results ============================================== The build and test results are cycled every 15 days. 2017-11-24 pass rate: 11558/11560, results: http://openjdk.linaro.org/jdkX/jcstress-nightly-runs/2017/327/results/ 2017-11-26 pass rate: 11558/11560, results: http://openjdk.linaro.org/jdkX/jcstress-nightly-runs/2017/329/results/ 2017-11-28 pass rate: 11559/11560, results: http://openjdk.linaro.org/jdkX/jcstress-nightly-runs/2017/331/results/ 2017-11-30 pass rate: 11557/11560, results: http://openjdk.linaro.org/jdkX/jcstress-nightly-runs/2017/333/results/ 2017-12-02 pass rate: 11557/11560, results: http://openjdk.linaro.org/jdkX/jcstress-nightly-runs/2017/335/results/ 2017-12-04 pass rate: 11557/11560, results: http://openjdk.linaro.org/jdkX/jcstress-nightly-runs/2017/337/results/ 2017-12-06 pass rate: 11558/11560, results: http://openjdk.linaro.org/jdkX/jcstress-nightly-runs/2017/339/results/ 2017-12-08 pass rate: 11556/11560, results: http://openjdk.linaro.org/jdkX/jcstress-nightly-runs/2017/341/results/ 2017-12-10 pass rate: 11559/11560, results: http://openjdk.linaro.org/jdkX/jcstress-nightly-runs/2017/343/results/ 2017-12-12 pass rate: 11558/11560, results: http://openjdk.linaro.org/jdkX/jcstress-nightly-runs/2017/345/results/ 2017-12-14 pass rate: 11558/11560, results: http://openjdk.linaro.org/jdkX/jcstress-nightly-runs/2017/347/results/ 2017-12-16 pass rate: 11557/11560, results: http://openjdk.linaro.org/jdkX/jcstress-nightly-runs/2017/349/results/ 2017-12-18 pass rate: 11558/11560, results: http://openjdk.linaro.org/jdkX/jcstress-nightly-runs/2017/351/results/ 2017-12-20 pass rate: 11559/11560, results: http://openjdk.linaro.org/jdkX/jcstress-nightly-runs/2017/353/results/ 2017-12-24 pass rate: 11557/11560, results: http://openjdk.linaro.org/jdkX/jcstress-nightly-runs/2017/357/results/ For detailed information on the test output please refer to: http://openjdk.linaro.org/jdkX/jcstress-nightly-runs/

1 0

[10] RFR: 8193699: AArch64 Fails to build after 8167372
by Rahul Raghavan 22 Dec '17

22 Dec '17

Hi, Request help to review the following fix proposal. <webrev> - http://cr.openjdk.java.net/~rraghavan/8193699/webrev.01/ Related <jbs> links - -- https://bugs.openjdk.java.net/browse/JDK-8193699 -- https://bugs.openjdk.java.net/browse/JDK-8167372 - 'Add code to check for getting oops while thread is in native' -- https://bugs.openjdk.java.net/browse/JDK-8191227 - 'Unsafe handle resolution in ConstantOopWriteValue::write_on() / print_on() and LIR_Assembler::jobject2reg()' Found root cause of 8193699 issue, as missing changes for aarch64 in 'macroAssembler_aarch64.cpp', similar to changes done for 8191227 before pushing 8167372 support. So now fixed the issue of unsafe handle resolution in 'macroAssembler_aarch64.cpp' with the usage of ThreadInVMfromUnknown. Stuart Monteith confirmed build failure fixed with above proposed changeset and no new issues with testing. Thanks, Rahul

3 4

JTREG, JCStress, SPECjbb2015 and Hadoop/Terasort results for OpenJDK JDK on AArch64
by ci_notify＠linaro.org 20 Dec '17

20 Dec '17

This is a summary of the JTREG test results =========================================== The build and test results are cycled every 15 days. For detailed information on the test output please refer to: http://openjdk.linaro.org/jdkX/openjdk-jtreg-nightly-tests/summary/2017/353… ------------------------------------------------------------------------------- client-release/hotspot ------------------------------------------------------------------------------- Build 0: aarch64/2017/nov/21 pass: 1,445; fail: 15 Build 1: aarch64/2017/nov/23 pass: 1,448; fail: 12 Build 2: aarch64/2017/nov/25 pass: 1,456; fail: 12; error: 1 Build 3: aarch64/2017/nov/27 pass: 1,457; fail: 11; error: 1 Build 4: aarch64/2017/nov/29 pass: 1,546; fail: 14 Build 5: aarch64/2017/dec/01 pass: 1,575; fail: 11 Build 6: aarch64/2017/dec/03 pass: 1,580; fail: 11 Build 7: aarch64/2017/dec/05 pass: 1,580; fail: 11 Build 8: aarch64/2017/dec/07 pass: 1,578; fail: 12; error: 1 Build 9: aarch64/2017/dec/09 pass: 1,584; fail: 10 Build 10: aarch64/2017/dec/11 pass: 1,584; fail: 10 Build 11: aarch64/2017/dec/13 pass: 1,585; fail: 9 Build 12: aarch64/2017/dec/15 pass: 1,588; fail: 10 Build 13: aarch64/2017/dec/17 pass: 1,597; fail: 11; error: 1 Build 14: aarch64/2017/dec/19 pass: 1,599; fail: 11 1 fatal errors were detected; please follow the link above for more detail. ------------------------------------------------------------------------------- client-release/jdk ------------------------------------------------------------------------------- Build 0: aarch64/2017/nov/21 pass: 7,490; fail: 739; error: 22 Build 1: aarch64/2017/nov/23 pass: 7,505; fail: 728; error: 22 Build 2: aarch64/2017/nov/25 pass: 7,531; fail: 697; error: 27 Build 3: aarch64/2017/nov/27 pass: 7,507; fail: 730; error: 18 Build 4: aarch64/2017/nov/29 pass: 7,501; fail: 737; error: 18 Build 5: aarch64/2017/dec/01 pass: 7,501; fail: 735; error: 20 Build 6: aarch64/2017/dec/03 pass: 7,532; fail: 710; error: 23 Build 7: aarch64/2017/dec/05 pass: 7,510; fail: 732; error: 23 Build 8: aarch64/2017/dec/07 pass: 7,527; fail: 715; error: 23 Build 9: aarch64/2017/dec/09 pass: 7,558; fail: 727; error: 22 Build 10: aarch64/2017/dec/11 pass: 7,559; fail: 732; error: 19 Build 11: aarch64/2017/dec/13 pass: 7,558; fail: 725; error: 27 Build 12: aarch64/2017/dec/15 pass: 7,569; fail: 741; error: 25 Build 13: aarch64/2017/dec/17 pass: 7,591; fail: 719; error: 26 Build 14: aarch64/2017/dec/19 pass: 7,572; fail: 740; error: 24 ------------------------------------------------------------------------------- client-release/langtools ------------------------------------------------------------------------------- Build 0: aarch64/2017/nov/21 pass: 3,812; fail: 4; error: 1 Build 1: aarch64/2017/nov/23 pass: 3,817; fail: 5; error: 1 Build 2: aarch64/2017/nov/25 pass: 3,818; fail: 4; error: 1 Build 3: aarch64/2017/nov/27 pass: 3,814; fail: 5; error: 4 Build 4: aarch64/2017/nov/29 pass: 3,820; fail: 3 Build 5: aarch64/2017/dec/01 pass: 3,818; fail: 3; error: 2 Build 6: aarch64/2017/dec/03 pass: 3,829; fail: 4; error: 1 Build 7: aarch64/2017/dec/05 pass: 3,828; fail: 4; error: 2 Build 8: aarch64/2017/dec/07 pass: 3,828; fail: 3; error: 3 Build 9: aarch64/2017/dec/09 pass: 3,830; fail: 4; error: 2 Build 10: aarch64/2017/dec/11 pass: 3,830; fail: 4; error: 1 Build 11: aarch64/2017/dec/13 pass: 3,831; fail: 4 Build 12: aarch64/2017/dec/15 pass: 3,835; fail: 3; error: 3 Build 13: aarch64/2017/dec/17 pass: 3,820; fail: 3; error: 3 Build 14: aarch64/2017/dec/19 pass: 3,819; fail: 3; error: 4 ------------------------------------------------------------------------------- server-release/hotspot ------------------------------------------------------------------------------- Build 0: aarch64/2017/nov/21 pass: 1,449; fail: 16 Build 1: aarch64/2017/nov/23 pass: 1,452; fail: 13 Build 2: aarch64/2017/nov/25 pass: 1,460; fail: 13; error: 1 Build 3: aarch64/2017/nov/27 pass: 1,460; fail: 13; error: 1 Build 4: aarch64/2017/nov/29 pass: 1,550; fail: 15 Build 5: aarch64/2017/dec/01 pass: 1,580; fail: 11 Build 6: aarch64/2017/dec/03 pass: 1,584; fail: 11; error: 1 Build 7: aarch64/2017/dec/05 pass: 1,585; fail: 10; error: 1 Build 8: aarch64/2017/dec/07 pass: 1,584; fail: 11; error: 1 Build 9: aarch64/2017/dec/09 pass: 1,588; fail: 10; error: 1 Build 10: aarch64/2017/dec/11 pass: 1,589; fail: 9; error: 1 Build 11: aarch64/2017/dec/13 pass: 1,589; fail: 9; error: 1 Build 12: aarch64/2017/dec/15 pass: 1,592; fail: 11 Build 13: aarch64/2017/dec/17 pass: 1,604; fail: 10 Build 14: aarch64/2017/dec/19 pass: 1,605; fail: 10 1 fatal errors were detected; please follow the link above for more detail. ------------------------------------------------------------------------------- server-release/jdk ------------------------------------------------------------------------------- Build 0: aarch64/2017/nov/21 pass: 7,508; fail: 721; error: 22 Build 1: aarch64/2017/nov/23 pass: 7,502; fail: 728; error: 25 Build 2: aarch64/2017/nov/25 pass: 7,511; fail: 715; error: 29 Build 3: aarch64/2017/nov/27 pass: 7,528; fail: 705; error: 22 Build 4: aarch64/2017/nov/29 pass: 7,512; fail: 719; error: 25 Build 5: aarch64/2017/dec/01 pass: 7,526; fail: 708; error: 22 Build 6: aarch64/2017/dec/03 pass: 7,530; fail: 710; error: 25 Build 7: aarch64/2017/dec/05 pass: 7,529; fail: 716; error: 20 Build 8: aarch64/2017/dec/07 pass: 7,543; fail: 690; error: 32 Build 9: aarch64/2017/dec/09 pass: 7,570; fail: 716; error: 21 Build 10: aarch64/2017/dec/11 pass: 7,584; fail: 705; error: 21 Build 11: aarch64/2017/dec/13 pass: 7,577; fail: 713; error: 20 Build 12: aarch64/2017/dec/15 pass: 7,600; fail: 711; error: 24 Build 13: aarch64/2017/dec/17 pass: 7,593; fail: 721; error: 22 Build 14: aarch64/2017/dec/19 pass: 7,586; fail: 726; error: 24 ------------------------------------------------------------------------------- server-release/langtools ------------------------------------------------------------------------------- Build 0: aarch64/2017/nov/21 pass: 3,812; fail: 4; error: 1 Build 1: aarch64/2017/nov/23 pass: 3,816; fail: 4; error: 3 Build 2: aarch64/2017/nov/25 pass: 3,819; fail: 4 Build 3: aarch64/2017/nov/27 pass: 3,819; fail: 3; error: 1 Build 4: aarch64/2017/nov/29 pass: 3,819; fail: 3; error: 1 Build 5: aarch64/2017/dec/01 pass: 3,819; fail: 4 Build 6: aarch64/2017/dec/03 pass: 3,829; fail: 3; error: 2 Build 7: aarch64/2017/dec/05 pass: 3,829; fail: 4; error: 1 Build 8: aarch64/2017/dec/07 pass: 3,829; fail: 3; error: 2 Build 9: aarch64/2017/dec/09 pass: 3,832; fail: 3; error: 1 Build 10: aarch64/2017/dec/11 pass: 3,830; fail: 3; error: 2 Build 11: aarch64/2017/dec/13 pass: 3,830; fail: 4; error: 1 Build 12: aarch64/2017/dec/15 pass: 3,833; fail: 3; error: 5 Build 13: aarch64/2017/dec/17 pass: 3,822; fail: 3; error: 1 Build 14: aarch64/2017/dec/19 pass: 3,817; fail: 3; error: 6 Previous results can be found here: http://openjdk.linaro.org/jdkX/openjdk-jtreg-nightly-tests/index.html SPECjbb2015 composite regression test completed =============================================== This test measures the relative performance of the server compiler running the SPECjbb2015 composite tests and compares the performance against the baseline performance of the server compiler taken on 2016-11-21. In accordance with [1], the SPECjbb2015 tests are run on a system which is not production ready and does not meet all the requirements for publishing compliant results. The numbers below shall be treated as non-compliant (nc) and are for experimental purposes only. Relative performance: Server max-jOPS (nc): 1.01x Relative performance: Server critical-jOPS (nc): 0.74x Details of the test setup and historical results may be found here: http://openjdk.linaro.org/jdkX/SPECjbb2015-results/ [1] http://www.spec.org/fairuse.html#Academic Regression test Hadoop-Terasort completed ========================================= This test measures the performance of the server and client compilers running Hadoop sorting a 1GB file using Terasort and compares the performance against the baseline performance of the Zero interpreter and against the baseline performance of the client and server compilers on 2014-04-01. Relative performance: Zero: 1.0, Client: 69.19, Server: 109.42 Client 69.19 / Client 2014-04-01 (43.00): 1.61x Server 109.42 / Server 2014-04-01 (71.00): 1.54x Details of the test setup and historical results may be found here: http://openjdk.linaro.org/jdkX/hadoop-terasort-benchmark-results/ This is a summary of the jcstress test results ============================================== The build and test results are cycled every 15 days. 2017-11-22 pass rate: 11558/11560, results: http://openjdk.linaro.org/jdkX/jcstress-nightly-runs/2017/325/results/ 2017-11-24 pass rate: 11558/11560, results: http://openjdk.linaro.org/jdkX/jcstress-nightly-runs/2017/327/results/ 2017-11-26 pass rate: 11558/11560, results: http://openjdk.linaro.org/jdkX/jcstress-nightly-runs/2017/329/results/ 2017-11-28 pass rate: 11559/11560, results: http://openjdk.linaro.org/jdkX/jcstress-nightly-runs/2017/331/results/ 2017-11-30 pass rate: 11557/11560, results: http://openjdk.linaro.org/jdkX/jcstress-nightly-runs/2017/333/results/ 2017-12-02 pass rate: 11557/11560, results: http://openjdk.linaro.org/jdkX/jcstress-nightly-runs/2017/335/results/ 2017-12-04 pass rate: 11557/11560, results: http://openjdk.linaro.org/jdkX/jcstress-nightly-runs/2017/337/results/ 2017-12-06 pass rate: 11558/11560, results: http://openjdk.linaro.org/jdkX/jcstress-nightly-runs/2017/339/results/ 2017-12-08 pass rate: 11556/11560, results: http://openjdk.linaro.org/jdkX/jcstress-nightly-runs/2017/341/results/ 2017-12-10 pass rate: 11559/11560, results: http://openjdk.linaro.org/jdkX/jcstress-nightly-runs/2017/343/results/ 2017-12-12 pass rate: 11558/11560, results: http://openjdk.linaro.org/jdkX/jcstress-nightly-runs/2017/345/results/ 2017-12-14 pass rate: 11558/11560, results: http://openjdk.linaro.org/jdkX/jcstress-nightly-runs/2017/347/results/ 2017-12-16 pass rate: 11557/11560, results: http://openjdk.linaro.org/jdkX/jcstress-nightly-runs/2017/349/results/ 2017-12-18 pass rate: 11558/11560, results: http://openjdk.linaro.org/jdkX/jcstress-nightly-runs/2017/351/results/ 2017-12-20 pass rate: 11559/11560, results: http://openjdk.linaro.org/jdkX/jcstress-nightly-runs/2017/353/results/ For detailed information on the test output please refer to: http://openjdk.linaro.org/jdkX/jcstress-nightly-runs/

1 0

Vectorized support for array equals/compare/mismatch using Unsafe
by Andrew Haley 19 Dec '17

19 Dec '17

https://bugs.openjdk.java.net/browse/JDK-8136924 I can't find a tracker bug for an AArch64 version of this. Is there one? -- Andrew Haley Java Platform Lead Engineer Red Hat UK Ltd. <https://www.redhat.com> EAC8 43EB D3EF DB98 CC77 2FAD A5CD 6035 332F A671

2 1