RFR[M]: 8151779: Some intrinsic flags could be replaced with one general flag

Vladimir Kozlov vladimir.kozlov at oracle.com
Sat May 2 03:29:49 UTC 2020


Hi Xin

compiler/intrinsics/IntrinsicAvailableTest.java test failed on old x86 machine which does not have CLMUL and as result 
can't use CRC32 intrinsic (UseCRC32Intrinsics flag is false). With  -XX:ControlIntrinsic=+_updateCRC32 test failed with:

java.lang.RuntimeException: Unexpected result: intrinsic for java.util.zip.CRC32.update() is enabled but intrinsic is 
not available at compilation level 1
	at compiler.intrinsics.IntrinsicAvailableTest.checkIntrinsicForCompilationLevel(IntrinsicAvailableTest.java:128)
	at compiler.intrinsics.IntrinsicAvailableTest.test(IntrinsicAvailableTest.java:138)
	at compiler.intrinsics.IntrinsicAvailableTest.main(IntrinsicAvailableTest.java:150)

Regards,
Vladimir

On 5/1/20 6:00 PM, Vladimir Kozlov wrote:
> Hi
> 
> I am CCing to runtime group too. I would like to see comments about these changes. No need to look on compiler's changes.
> 
> The latest https://cr.openjdk.java.net/~xliu/8151779/02/webrev/
> 
> Good work.
> 
> On 4/24/20 1:33 AM, Liu, Xin wrote:
>> Hi,
>>
>> May I get reviewed for this new revision?
>> JBS: https://bugs.openjdk.java.net/browse/JDK-8151779
>> webrev: https://cr.openjdk.java.net/~xliu/8151779/01/webrev/
>>
>> I introduce a new option -XX:ControlIntrinsic=+_id1,-id2...
>> The id is vmIntrinsics::ID.  As prior discussion, ControlIntrinsic is expected to replace DisableIntrinsic.
>> I keep DisableIntrinsic in this revision. DisableIntrinsic prevails when an intrinsic appears on both lists.
> 
> Yes, you have to keep DisableIntrinsic for now. We will deprecate it later.
> 
>>
>> I use an array of tribool to mark each intrinsic is enabled or not. In this way, hotspot can avoid expensive string 
>> querying among intrinsics.
>> A Tribool value has 3 states: Default, true, or false.
>> If developers don't explicitly set an intrinsic, it will be available unless is disabled by the corresponding 
>> UseXXXIntrinsics.
>> Traditional Boolean value can't express fine/coarse-grained control. Ie. We only go through those auxiliary options 
>> UseXXXIntrinsics if developers don't control a specific intrinsic.
>>
>> I also add the support of ControlIntrinsic to CompilerDirectives.
>>
>> Test:
>> I reuse jtreg tests of DisableIntrinsic. Add add more @run annotations to verify ControlIntrinsics.
>> I passed hotspot:Tier1 test and all tests on x86_64/linux.
> 
> Good. I submitted hotspot tier1-3 testing.
> 
> Thanks,
> Vladimir
> 
>>
>> Thanks,
>> --lx
>>
>> On 4/17/20, 7:22 PM, "hotspot-compiler-dev on behalf of Liu, Xin" <hotspot-compiler-dev-bounces at openjdk.java.net on 
>> behalf of xxinliu at amazon.com> wrote:
>>
>>      Hi, Vladimir,
>>
>>      Thanks for the clarification.
>>      Oh, yes, it's theoretically possible, but it's tedious. I am wrong at that point.
>>      I think I got your point. ControlIntrinsics will make developer's life easier. I will implement it.
>>
>>      Thanks,
>>      --lx
>>
>>
>>      On 4/17/20, 6:46 PM, "Vladimir Kozlov" <vladimir.kozlov at oracle.com> wrote:
>>
>>          CAUTION: This email originated from outside of the organization. Do not click links or open attachments 
>> unless you can confirm the sender and know the content is safe.
>>
>>
>>
>>          I withdraw my suggestion about EnableIntrinsic from JDK-8151779 because ControlIntrinsics will provide such
>>          functionality and will replace existing DisableIntrinsic.
>>
>>          Note, we can start deprecating Use*Intrinsic flags (and DisableIntrinsic) later in other changes. You don't 
>> need to do
>>          everything at once. What we need now a mechanism to replace them.
>>
>>          On 4/16/20 11:58 PM, Liu, Xin wrote:
>>          > Hi, Corey and Vladimir,
>>          >
>>          > I recently go through vmSymbols.hpp/cpp. I think I understand your comments.
>>          > Each UseXXXIntrinsics does control a bunch of intrinsics (plural). Thanks for the hint.
>>          >
>>          > Even though I feel I know intrinsics mechanism of hotspot better, I still need a clarification of JDK- 
>> 8151779.
>>          >
>>          > There're 321 intrinsics (https://chriswhocodes.com/hotspot_intrinsics_jdk15.html).
>>          > If there's no any option, they are all available for compilers.  That makes sense because intrinsics are 
>> always beneficial.
>>          > But there're reasons we need to disable a subset of them. A specific architecture may miss efficient 
>> instructions or fixed functions. Or simply because an intrinsic is buggy.
>>          >
>>          > Currently, JDK provides developers 2 ways to control intrinsics. > 1. Some diagnostic options. Eg. 
>> InlineMathNatives, UseBase64Intrinsics.
>>          > Developers can use one option to disable a group of intrinsics.  That is to say, it's a coarse-grained 
>> approach.
>>          >
>>          > 2. DisableIntrinsic="a,b,c"
>>          > By passing a string list of vmIntrinsics::IDs, it's capable of disabling any specified intrinsic.
>>          >
>>          > But even putting above 2 approaches together, we still can't precisely control any intrinsic.
>>
>>          Yes, you are right. We seems are trying to put these 2 different ways into one flag which may be mistake.
>>
>>          -XX:ControlIntrinsic=-_updateBytesCRC32C,-_updateDirectByteBufferCRC32C is a similar to 
>> -XX:-UseCRC32CIntrinsics but it
>>          requires more detailed knowledge about intrinsics ids.
>>
>>          May be we can have 2nd flag, as you suggested -XX:UseIntrinsics=-AESCTR,+CRC32C, for such cases.
>>
>>          > If we want to enable an intrinsic which is under control of InlineMathNatives but keep others disable, it's 
>> impossible now.  [please correct if I am wrong here].
>>
>>          You can disable all other from 321 intrinsics with DisableIntrinsic flag which is very tedious I agree.
>>
>>          > I think that the motivation JDK-8151779 tried to solve.
>>
>>          The idea is that instead of flags we use to control particular intrinsics depending on CPU we will use 
>> vmIntrinsics::IDs
>>          or other tables as you showed in your changes. It will require changes in vm_version_<cpu> codes.
>>
>>          >
>>          > If we provide a new option EnableIntrinsic and put it least priority, then we can precisely control any 
>> intrinsic.
>>          > Quote Vladimir Kozlov "DisableIntrinsic list prevails if an intrinsic is specified on both EnableIntrinsic 
>> and DisableIntrinsic."
>>          >
>>          >   "-XX:ControlIntrinsic=+_dabs,-_fabs,-_getClass" looks more elegant, but it will confuse developers with 
>> DisableIntrinsic.
>>          > If we decide to deprecate DisableIntrinsic, I think ControlIntrinsic may be a better option. Now I prefer 
>> to provide EnableIntrinsic for simplicity and symmetry.
>>
>>          I prefer to have one ControlIntrinsic flag and deprecate DisableIntrinsic. I don't think it is confusing.
>>
>>          Thanks,
>>          Vladimir
>>
>>          > What do you think?
>>          >
>>          > Thanks,
>>          > --lx
>>          >
>>          >
>>          > On 4/13/20, 1:47 PM, "hotspot-compiler-dev on behalf of Corey Ashford" 
>> <hotspot-compiler-dev-bounces at openjdk.java.net on behalf of cjashfor at linux.ibm.com> wrote:
>>          >
>>          >      CAUTION: This email originated from outside of the organization. Do not click links or open 
>> attachments unless you can confirm the sender and know the content is safe.
>>          >
>>          >
>>          >
>>          >      On 4/13/20 10:33 AM, Liu, Xin wrote:
>>          >      > Hi, compiler developers,
>>          >      > I attempt to refactor UseXXXIntrinsics for JDK-8151779.  I think we still need to keep 
>> UseXXXIntrinsics options because many applications may be using them.
>>          >      >
>>          >      > My change provide 2 new features:
>>          >      > 1) a shorthand to enable/disable intrinsics.
>>          >      > A comma-separated string. Each one is an intrinsic. An optional tailing symbol + or '-' denotes 
>> enabling or disabling.
>>          >      > If the tailing symbol is missing, it means enable.
>>          >      > Eg. -XX:UseIntrinsics="AESCTR-,CRC32C+,CRC32-,MathExact"
>>          >      > This jvm option will expand to multiple options -XX:-UseAESCTRIntrinsics, -XX:+UseCRC32CIntrinsics, 
>> -XX:-UseCRC32Intrinsics, -XX:UseMathExactIntrinsics
>>          >      >
>>          >      > 2) provide a set of macro to declare intrinsic options
>>          >      > Developers declare once in intrinsics.hpp and macros will take care all other places.
>>          >      > Here are example: 
>> https://cr.openjdk.java.net/~xliu/8151779/00/webrev/src/hotspot/share/compiler/intrinsics.hpp.html
>>          >      > Ion Lam is overhauling jvm options.  I am thinking how to be consistent with his proposal.
>>          >      >
>>          >
>>          >      Great idea, though to be consistent with the original syntax, I think
>>          >      the +/- should be in front of the name:
>>          >
>>          >      -XX:UseIntrinsics=-AESCTR,+CRC32C,...
>>          >
>>          >
>>          >      > I handle UseIntrinsics before VM_Version::initialize. It means that platform-specific initialization 
>> still has a chance to correct those options.
>>          >      > If we do that after VM_Version::initialize,  some intrinsics may cause JVM crash.  Eg. 
>> +UseBase64Intrinsics on x86_64 Linux.
>>          >      > Even though this behavior is same as -XX:+UseXXXIntrinsics, from user's perspective, it's not 
>> straightforward when JVM overrides what users specify implicitly. It's dilemma here,  stable jvm or fidelity of 
>> cmdline.  What do you think?
>>          >      >
>>          >      > Another problem is naming convention. Almost all intrinsics options use UseXXXIntrinsics. One 
>> exception is UseVectorizedMismatchIntrinsic.
>>          >      > Personally, I think it should be "UseXXXIntrinsic" because one option is for one intrinsic, right?  
>> Is it possible to change this name convention?
>>          >
>>          >      Some (many?) intrinsic options turn on more than one .ad instruct
>>          >      instrinsic, or library instrinsics at the same time.  I think that's why
>>          >      the plural is there.  Also, consistently adding the plural allows you to
>>          >      add more capabilities to a flag that initially only had one intrinsic
>>          >      without changing the plurality (and thus backward compatibility).
>>          >
>>          >      Regards,
>>          >
>>          >      - Corey
>>          >
>>          >
>>
>>


More information about the hotspot-compiler-dev mailing list