From duke at openjdk.java.net  Tue Feb  1 11:12:40 2022
From: duke at openjdk.java.net (Aggelos Biboudis)
Date: Tue, 1 Feb 2022 11:12:40 GMT
Subject: [lworld] RFR: 8280194: Abstract classes that allow value
 subclasses should be marked ACC_PERMITS_VALUE (0x0040)
In-Reply-To: <6R7uyjsabNXiT2YljFGI3ohOzBDtOTIYMF-MdK6tncI=.4ff4eb00-4cdc-4123-887c-a827cc4dc35b@github.com>
References: <6R7uyjsabNXiT2YljFGI3ohOzBDtOTIYMF-MdK6tncI=.4ff4eb00-4cdc-4123-887c-a827cc4dc35b@github.com>
Message-ID: <5r-ovi2Ti16W-v8u25-U9-H0w26_DsybBJGWXh6-XWc=.a48f53a6-f064-4b3d-9e37-d5d0afdcb732@github.com>

On Mon, 31 Jan 2022 14:08:59 GMT, Srikanth Adayapalam <sadayapalam at openjdk.org> wrote:

> Mark abstract classes that qualify with ACC_PERMITS_VALUE flag.

This PR looks good. Some clarifications requested.

-------------

Marked as reviewed by biboudis at github.com (no known OpenJDK username).

PR: https://git.openjdk.java.net/valhalla/pull/622

From duke at openjdk.java.net  Tue Feb  1 11:12:44 2022
From: duke at openjdk.java.net (Aggelos Biboudis)
Date: Tue, 1 Feb 2022 11:12:44 GMT
Subject: [lworld] RFR: 8280194: Abstract classes that allow value
 subclasses should be marked ACC_PERMITS_VALUE (0x0040)
In-Reply-To: <sxbvEDn5_VkzM_gPwjfJhEk285Saty-AAcnl1KJzqjg=.dbb8f56f-bc8c-4f15-9f05-acf9ea757c42@github.com>
References: <6R7uyjsabNXiT2YljFGI3ohOzBDtOTIYMF-MdK6tncI=.4ff4eb00-4cdc-4123-887c-a827cc4dc35b@github.com>
 <sxbvEDn5_VkzM_gPwjfJhEk285Saty-AAcnl1KJzqjg=.dbb8f56f-bc8c-4f15-9f05-acf9ea757c42@github.com>
Message-ID: <GFrv-KBt2VvX0fSsogct63g52qIDQoybmzd7XdPjmGk=.a3efbd79-a3e5-42cf-b591-4d359c2f30de@github.com>

On Mon, 31 Jan 2022 14:10:03 GMT, Srikanth Adayapalam <sadayapalam at openjdk.org> wrote:

>> Mark abstract classes that qualify with ACC_PERMITS_VALUE flag.
>
> src/jdk.compiler/share/classes/com/sun/tools/javac/code/Types.java line 2381:
> 
>> 2379:                             if ((s.flags() & (SYNCHRONIZED | STATIC)) == SYNCHRONIZED) {
>> 2380:                                 return true;
>> 2381:                             } else if (s.isConstructor()) {
> 
> This is a bug fix - static methods may be synchronized or not - immaterial.

Do you think the comment above the signature of `implicitIdentityType` needs to reflect this change in a parenthesis?

> src/jdk.compiler/share/classes/com/sun/tools/javac/comp/Attr.java line 1255:
> 
>> 1253:                     }
>> 1254:                 }
>> 1255: 
> 
> This is too late here. Now this is moved up to MemberEnter

Why was this moved up? (for me to understand better)

> test/langtools/tools/javac/classfiles/attributes/innerclasses/InnerClassesInLocalClassTest.java line 56:
> 
>> 54:             sb.append(outerModifier.getString()).append(' ');
>> 55:             sb.append("class Local { int f; ");
>> 56:             Map<String, Set<String>> class2Flags = new HashMap<>();
> 
> Hack to suppress ACC_PERMITS_VALUE bit from being set thereby causing the test to "fail"

I was playing around to understand this generated test:


public class OuterClass {
 static void method() {abstract class Local { class A0{}
abstract class A1{}
final class A2{}
};}
}


Why do we get the following for an inner class that is abstract,


[ASSERT] : inner_class_access_flags 1Local
Expected: [ACC_ABSTRACT],
     Got: [ACC_ABSTRACT, 0x40]


while `abstract class A8` in `PermitsValueTest.java` does not emit `ACC_PERMITS_VALUE`?

nitpick (and possibly not addressed yet): `Modifiers:getString()` doesn't return the name for `ACC_PERMITS_VALUE` yet. Do we need to address this in this PR? possibly not

-------------

PR: https://git.openjdk.java.net/valhalla/pull/622

From sadayapalam at openjdk.java.net  Tue Feb  1 11:20:37 2022
From: sadayapalam at openjdk.java.net (Srikanth Adayapalam)
Date: Tue, 1 Feb 2022 11:20:37 GMT
Subject: [lworld] RFR: 8280194: Abstract classes that allow value
 subclasses should be marked ACC_PERMITS_VALUE (0x0040)
In-Reply-To: <GFrv-KBt2VvX0fSsogct63g52qIDQoybmzd7XdPjmGk=.a3efbd79-a3e5-42cf-b591-4d359c2f30de@github.com>
References: <6R7uyjsabNXiT2YljFGI3ohOzBDtOTIYMF-MdK6tncI=.4ff4eb00-4cdc-4123-887c-a827cc4dc35b@github.com>
 <sxbvEDn5_VkzM_gPwjfJhEk285Saty-AAcnl1KJzqjg=.dbb8f56f-bc8c-4f15-9f05-acf9ea757c42@github.com>
 <GFrv-KBt2VvX0fSsogct63g52qIDQoybmzd7XdPjmGk=.a3efbd79-a3e5-42cf-b591-4d359c2f30de@github.com>
Message-ID: <XycCTj-yIueSGxKcXKYCmU6kXFHqCO7nnP_x5cOMKF8=.eb0617de-c521-49f3-a840-9911302e8377@github.com>

On Tue, 1 Feb 2022 10:08:18 GMT, Aggelos Biboudis <duke at openjdk.java.net> wrote:

> Do you think the comment above the signature of `implicitIdentityType` needs to reflect this change in a parenthesis?

Sure, I don't see a comment above implicitIdentityType, but inside it at the top, I have amended it to say "if it declares a synhronized instance method ..."

-------------

PR: https://git.openjdk.java.net/valhalla/pull/622

From sadayapalam at openjdk.java.net  Tue Feb  1 11:25:40 2022
From: sadayapalam at openjdk.java.net (Srikanth Adayapalam)
Date: Tue, 1 Feb 2022 11:25:40 GMT
Subject: [lworld] RFR: 8280194: Abstract classes that allow value
 subclasses should be marked ACC_PERMITS_VALUE (0x0040)
In-Reply-To: <GFrv-KBt2VvX0fSsogct63g52qIDQoybmzd7XdPjmGk=.a3efbd79-a3e5-42cf-b591-4d359c2f30de@github.com>
References: <6R7uyjsabNXiT2YljFGI3ohOzBDtOTIYMF-MdK6tncI=.4ff4eb00-4cdc-4123-887c-a827cc4dc35b@github.com>
 <sxbvEDn5_VkzM_gPwjfJhEk285Saty-AAcnl1KJzqjg=.dbb8f56f-bc8c-4f15-9f05-acf9ea757c42@github.com>
 <GFrv-KBt2VvX0fSsogct63g52qIDQoybmzd7XdPjmGk=.a3efbd79-a3e5-42cf-b591-4d359c2f30de@github.com>
Message-ID: <WEGzs8vPtOTpbccgmn_IHzAjTH2SLs85Cc_ZqhH3JLE=.97b40244-e529-42cb-b9c4-3570c4dc859c@github.com>

On Tue, 1 Feb 2022 10:10:02 GMT, Aggelos Biboudis <duke at openjdk.java.net> wrote:

>> src/jdk.compiler/share/classes/com/sun/tools/javac/comp/Attr.java line 1255:
>> 
>>> 1253:                     }
>>> 1254:                 }
>>> 1255: 
>> 
>> This is too late here. Now this is moved up to MemberEnter
>
> Why was this moved up? (for me to understand better)

This was in being set in Attribution phase - and is also consulted in the attribution phase. 
so when we have classes like:

value class V extends A {}
abstract class A { A() {} } 

when we are attributing V, A may not be attributed. While attributing V, we will check if V's superclass (A) meets all the requirements of an abstract class to be a value super but A isn't attributed yet.

What we have done now is to move these analysis to MemberEnter - MemberEnter for V and A are both guaranteed to have finished before Attribution of either starts, so we have eliminated the phase ordering problem

-------------

PR: https://git.openjdk.java.net/valhalla/pull/622

From sadayapalam at openjdk.java.net  Tue Feb  1 11:33:23 2022
From: sadayapalam at openjdk.java.net (Srikanth Adayapalam)
Date: Tue, 1 Feb 2022 11:33:23 GMT
Subject: [lworld] RFR: 8280194: Abstract classes that allow value
 subclasses should be marked ACC_PERMITS_VALUE (0x0040) [v2]
In-Reply-To: <6R7uyjsabNXiT2YljFGI3ohOzBDtOTIYMF-MdK6tncI=.4ff4eb00-4cdc-4123-887c-a827cc4dc35b@github.com>
References: <6R7uyjsabNXiT2YljFGI3ohOzBDtOTIYMF-MdK6tncI=.4ff4eb00-4cdc-4123-887c-a827cc4dc35b@github.com>
Message-ID: <Q4ghm3RJlgzwFYN9we3znt5tSdwnA2o79qbEy-v0Fv4=.49d0e22d-70c7-4287-b29c-46859088ec65@github.com>

> Mark abstract classes that qualify with ACC_PERMITS_VALUE flag.

Srikanth Adayapalam has updated the pull request incrementally with one additional commit since the last revision:

  Incorporate review comments

-------------

Changes:
  - all: https://git.openjdk.java.net/valhalla/pull/622/files
  - new: https://git.openjdk.java.net/valhalla/pull/622/files/28ff0c54..7f41d9a4

Webrevs:
 - full: https://webrevs.openjdk.java.net/?repo=valhalla&pr=622&range=01
 - incr: https://webrevs.openjdk.java.net/?repo=valhalla&pr=622&range=00-01

  Stats: 1 line in 1 file changed: 0 ins; 0 del; 1 mod
  Patch: https://git.openjdk.java.net/valhalla/pull/622.diff
  Fetch: git fetch https://git.openjdk.java.net/valhalla pull/622/head:pull/622

PR: https://git.openjdk.java.net/valhalla/pull/622

From sadayapalam at openjdk.java.net  Tue Feb  1 11:33:23 2022
From: sadayapalam at openjdk.java.net (Srikanth Adayapalam)
Date: Tue, 1 Feb 2022 11:33:23 GMT
Subject: [lworld] RFR: 8280194: Abstract classes that allow value
 subclasses should be marked ACC_PERMITS_VALUE (0x0040) [v2]
In-Reply-To: <GFrv-KBt2VvX0fSsogct63g52qIDQoybmzd7XdPjmGk=.a3efbd79-a3e5-42cf-b591-4d359c2f30de@github.com>
References: <6R7uyjsabNXiT2YljFGI3ohOzBDtOTIYMF-MdK6tncI=.4ff4eb00-4cdc-4123-887c-a827cc4dc35b@github.com>
 <sxbvEDn5_VkzM_gPwjfJhEk285Saty-AAcnl1KJzqjg=.dbb8f56f-bc8c-4f15-9f05-acf9ea757c42@github.com>
 <GFrv-KBt2VvX0fSsogct63g52qIDQoybmzd7XdPjmGk=.a3efbd79-a3e5-42cf-b591-4d359c2f30de@github.com>
Message-ID: <CNz4eNRlmj-CTOfDd7Zu_r_HFr1q3_MZkM_-3yWkIkk=.51f7325b-7bbb-4c68-9d06-da3d8735241c@github.com>

On Tue, 1 Feb 2022 11:04:48 GMT, Aggelos Biboudis <duke at openjdk.java.net> wrote:

> I was playing around to understand this generated test:
> 
> ```
> public class OuterClass {
>  static void method() {abstract class Local { class A0{}
> abstract class A1{}
> final class A2{}
> };}
> }
> ```
> 
> Why do we get the following for an inner class that is abstract,
> 
> ```
> [ASSERT] : inner_class_access_flags 1Local
> Expected: [ACC_ABSTRACT],
>      Got: [ACC_ABSTRACT, 0x40]
> ```
> 
> while `abstract class A8` in `PermitsValueTest.java` does not emit `ACC_PERMITS_VALUE`?

Class A8 in PermitsValue is an inner class that has an enclosing instance. Enclosing instance means the class will have an instance field to store the outer this and also its constructor will receive an argument - ie it cannot be a no-arg constructor only. (You may want to see the class files generated to understand the synthetic fields and constructor args that represent the outer this for an inner class with an enclosing instance) 

> 
> nitpick (and possibly not addressed yet): `Modifiers:getString()` doesn't return the name for `ACC_PERMITS_VALUE` yet. Do we need to address this in this PR? possibly not

Oops - Yes I will raise a follow up defect.

-------------

PR: https://git.openjdk.java.net/valhalla/pull/622

From sadayapalam at openjdk.java.net  Tue Feb  1 11:37:41 2022
From: sadayapalam at openjdk.java.net (Srikanth Adayapalam)
Date: Tue, 1 Feb 2022 11:37:41 GMT
Subject: [lworld] Integrated: 8280194: Abstract classes that allow value
 subclasses should be marked ACC_PERMITS_VALUE (0x0040)
In-Reply-To: <6R7uyjsabNXiT2YljFGI3ohOzBDtOTIYMF-MdK6tncI=.4ff4eb00-4cdc-4123-887c-a827cc4dc35b@github.com>
References: <6R7uyjsabNXiT2YljFGI3ohOzBDtOTIYMF-MdK6tncI=.4ff4eb00-4cdc-4123-887c-a827cc4dc35b@github.com>
Message-ID: <om7_cNzO0bfXeebxYVFcgwl2cgn2WvAaxBLmXDuGK9Y=.964c3d30-bcce-4ea3-bc4c-aee50a5e2762@github.com>

On Mon, 31 Jan 2022 14:08:59 GMT, Srikanth Adayapalam <sadayapalam at openjdk.org> wrote:

> Mark abstract classes that qualify with ACC_PERMITS_VALUE flag.

This pull request has now been integrated.

Changeset: 2e44552e
Author:    Srikanth Adayapalam <sadayapalam at openjdk.org>
URL:       https://git.openjdk.java.net/valhalla/commit/2e44552e0b8566280a706cfbec73eaec1bffe6fb
Stats:     274 lines in 16 files changed: 228 ins; 10 del; 36 mod

8280194: Abstract classes that allow value subclasses should be marked ACC_PERMITS_VALUE (0x0040)
8242619: [lworld] Ascertain that no-arg super constructors of inline classes are empty

-------------

PR: https://git.openjdk.java.net/valhalla/pull/622

From duke at openjdk.java.net  Tue Feb  1 12:35:41 2022
From: duke at openjdk.java.net (Aggelos Biboudis)
Date: Tue, 1 Feb 2022 12:35:41 GMT
Subject: [lworld] RFR: 8280194: Abstract classes that allow value
 subclasses should be marked ACC_PERMITS_VALUE (0x0040) [v2]
In-Reply-To: <CNz4eNRlmj-CTOfDd7Zu_r_HFr1q3_MZkM_-3yWkIkk=.51f7325b-7bbb-4c68-9d06-da3d8735241c@github.com>
References: <6R7uyjsabNXiT2YljFGI3ohOzBDtOTIYMF-MdK6tncI=.4ff4eb00-4cdc-4123-887c-a827cc4dc35b@github.com>
 <sxbvEDn5_VkzM_gPwjfJhEk285Saty-AAcnl1KJzqjg=.dbb8f56f-bc8c-4f15-9f05-acf9ea757c42@github.com>
 <GFrv-KBt2VvX0fSsogct63g52qIDQoybmzd7XdPjmGk=.a3efbd79-a3e5-42cf-b591-4d359c2f30de@github.com>
 <CNz4eNRlmj-CTOfDd7Zu_r_HFr1q3_MZkM_-3yWkIkk=.51f7325b-7bbb-4c68-9d06-da3d8735241c@github.com>
Message-ID: <8oPO76EdvrFG_xgImynGZGiRHR3Bi4uWHte_SNy1YOc=.a96cc066-f862-4a5e-bff5-8f20c19c1b19@github.com>

On Tue, 1 Feb 2022 11:27:08 GMT, Srikanth Adayapalam <sadayapalam at openjdk.org> wrote:

>> I was playing around to understand this generated test:
>> 
>> 
>> public class OuterClass {
>>  static void method() {abstract class Local { class A0{}
>> abstract class A1{}
>> final class A2{}
>> };}
>> }
>> 
>> 
>> Why do we get the following for an inner class that is abstract,
>> 
>> 
>> [ASSERT] : inner_class_access_flags 1Local
>> Expected: [ACC_ABSTRACT],
>>      Got: [ACC_ABSTRACT, 0x40]
>> 
>> 
>> while `abstract class A8` in `PermitsValueTest.java` does not emit `ACC_PERMITS_VALUE`?
>> 
>> nitpick (and possibly not addressed yet): `Modifiers:getString()` doesn't return the name for `ACC_PERMITS_VALUE` yet. Do we need to address this in this PR? possibly not
>
>> I was playing around to understand this generated test:
>> 
>> ```
>> public class OuterClass {
>>  static void method() {abstract class Local { class A0{}
>> abstract class A1{}
>> final class A2{}
>> };}
>> }
>> ```
>> 
>> Why do we get the following for an inner class that is abstract,
>> 
>> ```
>> [ASSERT] : inner_class_access_flags 1Local
>> Expected: [ACC_ABSTRACT],
>>      Got: [ACC_ABSTRACT, 0x40]
>> ```
>> 
>> while `abstract class A8` in `PermitsValueTest.java` does not emit `ACC_PERMITS_VALUE`?
> 
> Class A8 in PermitsValue is an inner class that has an enclosing instance. Enclosing instance means the class will have an instance field to store the outer this and also its constructor will receive an argument - ie it cannot be a no-arg constructor only. (You may want to see the class files generated to understand the synthetic fields and constructor args that represent the outer this for an inner class with an enclosing instance) 
> 
>> 
>> nitpick (and possibly not addressed yet): `Modifiers:getString()` doesn't return the name for `ACC_PERMITS_VALUE` yet. Do we need to address this in this PR? possibly not
> 
> Oops - Yes I will raise a follow up defect.

Thank you! It's clear now!

-------------

PR: https://git.openjdk.java.net/valhalla/pull/622

From fparain at openjdk.java.net  Wed Feb  2 13:25:12 2022
From: fparain at openjdk.java.net (Frederic Parain)
Date: Wed, 2 Feb 2022 13:25:12 GMT
Subject: [lworld] RFR: 8281116: [lworld] Adding Preload attribute support
Message-ID: <CRTIZH_sHFnIXrvapvslAal1vhVNTsOdTNYBpJ-V7D0=.e2b4b5f2-1eae-4012-b266-e05e2133f477@github.com>

Please review this changeset adding support for the Preload attribute.
It includes a logging feature to track and diagnose those optimistic class loading.

Tested with Mach5 tier1.

Thank you,

Fred

-------------

Commit messages:
 - Fix Tabs
 - Adding support for the Preload attribute

Changes: https://git.openjdk.java.net/valhalla/pull/623/files
 Webrev: https://webrevs.openjdk.java.net/?repo=valhalla&pr=623&range=00
  Issue: https://bugs.openjdk.java.net/browse/JDK-8281116
  Stats: 376 lines in 10 files changed: 372 ins; 0 del; 4 mod
  Patch: https://git.openjdk.java.net/valhalla/pull/623.diff
  Fetch: git fetch https://git.openjdk.java.net/valhalla pull/623/head:pull/623

PR: https://git.openjdk.java.net/valhalla/pull/623

From hseigel at openjdk.java.net  Wed Feb  2 14:14:35 2022
From: hseigel at openjdk.java.net (Harold Seigel)
Date: Wed, 2 Feb 2022 14:14:35 GMT
Subject: [lworld] RFR: 8281116: [lworld] Adding Preload attribute support
In-Reply-To: <CRTIZH_sHFnIXrvapvslAal1vhVNTsOdTNYBpJ-V7D0=.e2b4b5f2-1eae-4012-b266-e05e2133f477@github.com>
References: <CRTIZH_sHFnIXrvapvslAal1vhVNTsOdTNYBpJ-V7D0=.e2b4b5f2-1eae-4012-b266-e05e2133f477@github.com>
Message-ID: <2C7cP_9svnP2WH4zLEtpKM3JBRqnCk32trLckrXbWvo=.54a04f9b-b024-4c47-b82a-5b13ec716f5f@github.com>

On Wed, 2 Feb 2022 13:12:19 GMT, Frederic Parain <fparain at openjdk.org> wrote:

> Please review this changeset adding support for the Preload attribute.
> It includes a logging feature to track and diagnose those optimistic class loading.
> 
> Tested with Mach5 tier1.
> 
> Thank you,
> 
> Fred

src/hotspot/share/oops/instanceKlass.hpp line 291:

> 289:     _misc_has_preload_attribute               = 1 << 23  // class has a Preload attribute
> 290:   };
> 291: 

Is the _misc_has_preload_attribute flag needed, or can the code just check if _preload_classes equals Null?

-------------

PR: https://git.openjdk.java.net/valhalla/pull/623

From fparain at openjdk.java.net  Wed Feb  2 14:21:33 2022
From: fparain at openjdk.java.net (Frederic Parain)
Date: Wed, 2 Feb 2022 14:21:33 GMT
Subject: [lworld] RFR: 8281116: [lworld] Adding Preload attribute support
In-Reply-To: <2C7cP_9svnP2WH4zLEtpKM3JBRqnCk32trLckrXbWvo=.54a04f9b-b024-4c47-b82a-5b13ec716f5f@github.com>
References: <CRTIZH_sHFnIXrvapvslAal1vhVNTsOdTNYBpJ-V7D0=.e2b4b5f2-1eae-4012-b266-e05e2133f477@github.com>
 <2C7cP_9svnP2WH4zLEtpKM3JBRqnCk32trLckrXbWvo=.54a04f9b-b024-4c47-b82a-5b13ec716f5f@github.com>
Message-ID: <0QsHysPcBVc0S0S0rZNaeQTZv5EjtDT8IX2UuUZrrIg=.65572a8d-938c-46ed-93ff-3f21818cd168@github.com>

On Wed, 2 Feb 2022 14:11:27 GMT, Harold Seigel <hseigel at openjdk.org> wrote:

>> Please review this changeset adding support for the Preload attribute.
>> It includes a logging feature to track and diagnose those optimistic class loading.
>> 
>> Tested with Mach5 tier1.
>> 
>> Thank you,
>> 
>> Fred
>
> src/hotspot/share/oops/instanceKlass.hpp line 291:
> 
>> 289:     _misc_has_preload_attribute               = 1 << 23  // class has a Preload attribute
>> 290:   };
>> 291: 
> 
> Is the _misc_has_preload_attribute flag needed, or can the code just check if _preload_classes equals Null?

Strictly speaking, it is not needed. At some point, I've considered putting the _preload_classes pointer into the optional part of InstanceKlass, which would have required this flag. But finally it is declared as a regular field, so this flag can be removed.

-------------

PR: https://git.openjdk.java.net/valhalla/pull/623

From hseigel at openjdk.java.net  Wed Feb  2 14:25:42 2022
From: hseigel at openjdk.java.net (Harold Seigel)
Date: Wed, 2 Feb 2022 14:25:42 GMT
Subject: [lworld] RFR: 8281116: [lworld] Adding Preload attribute support
In-Reply-To: <CRTIZH_sHFnIXrvapvslAal1vhVNTsOdTNYBpJ-V7D0=.e2b4b5f2-1eae-4012-b266-e05e2133f477@github.com>
References: <CRTIZH_sHFnIXrvapvslAal1vhVNTsOdTNYBpJ-V7D0=.e2b4b5f2-1eae-4012-b266-e05e2133f477@github.com>
Message-ID: <zlOl9xgtng_-Uhdkb_asJEK39E1SnTW-LfzBYos2e0s=.78e579c5-aba5-44a8-9634-6ccf2461c55e@github.com>

On Wed, 2 Feb 2022 13:12:19 GMT, Frederic Parain <fparain at openjdk.org> wrote:

> Please review this changeset adding support for the Preload attribute.
> It includes a logging feature to track and diagnose those optimistic class loading.
> 
> Tested with Mach5 tier1.
> 
> Thank you,
> 
> Fred

src/hotspot/share/classfile/classFileParser.cpp line 5951:

> 5949:   _record_components(NULL),
> 5950:   _temp_local_interfaces(NULL),
> 5951:   _local_interfaces(NULL),

Do you need to free _preload_classes in ClassFileParser::~ClassFileParser() and set _preload_classes to NULL in ClassFileParser::clear_class_metadata(), also, assert that it is NULL in ClassFileParser::fill_instance_klass() ?

-------------

PR: https://git.openjdk.java.net/valhalla/pull/623

From hseigel at openjdk.java.net  Wed Feb  2 14:29:41 2022
From: hseigel at openjdk.java.net (Harold Seigel)
Date: Wed, 2 Feb 2022 14:29:41 GMT
Subject: [lworld] RFR: 8281116: [lworld] Adding Preload attribute support
In-Reply-To: <CRTIZH_sHFnIXrvapvslAal1vhVNTsOdTNYBpJ-V7D0=.e2b4b5f2-1eae-4012-b266-e05e2133f477@github.com>
References: <CRTIZH_sHFnIXrvapvslAal1vhVNTsOdTNYBpJ-V7D0=.e2b4b5f2-1eae-4012-b266-e05e2133f477@github.com>
Message-ID: <VbhTzJcemDmphskQ0ECrRjwqtNXVylIokLNTMIzYlZY=.d4720b9c-5d1e-40ea-8215-62692d675e8f@github.com>

On Wed, 2 Feb 2022 13:12:19 GMT, Frederic Parain <fparain at openjdk.org> wrote:

> Please review this changeset adding support for the Preload attribute.
> It includes a logging feature to track and diagnose those optimistic class loading.
> 
> Tested with Mach5 tier1.
> 
> Thank you,
> 
> Fred

test/hotspot/jtreg/runtime/valhalla/inlinetypes/ValuePreloadTest.java line 61:

> 59:         checkFor(pb, "[warning][class,preload] Preloading of class PreloadValue1 during linking of class ValuePreloadClient1 (Preload attribute) failed");
> 60:     }
> 61: }

The test looks good but have multipe 2022's in the copyright:
 * Copyright (c) 2022, 2022, Oracle and/or its affiliates. All rights reserved.

-------------

PR: https://git.openjdk.java.net/valhalla/pull/623

From fparain at openjdk.java.net  Wed Feb  2 15:58:22 2022
From: fparain at openjdk.java.net (Frederic Parain)
Date: Wed, 2 Feb 2022 15:58:22 GMT
Subject: [lworld] RFR: 8281116: [lworld] Adding Preload attribute support
 [v2]
In-Reply-To: <CRTIZH_sHFnIXrvapvslAal1vhVNTsOdTNYBpJ-V7D0=.e2b4b5f2-1eae-4012-b266-e05e2133f477@github.com>
References: <CRTIZH_sHFnIXrvapvslAal1vhVNTsOdTNYBpJ-V7D0=.e2b4b5f2-1eae-4012-b266-e05e2133f477@github.com>
Message-ID: <IctDcuIUhv7s1dfaocc6sHdAatFYuoC7jrFKZ2JYNdw=.0a22c9aa-20da-46e3-8bb7-60ca10997f03@github.com>

> Please review this changeset adding support for the Preload attribute.
> It includes a logging feature to track and diagnose those optimistic class loading.
> 
> Tested with Mach5 tier1.
> 
> Thank you,
> 
> Fred

Frederic Parain has updated the pull request incrementally with one additional commit since the last revision:

  Fixes to address Harold's comments

-------------

Changes:
  - all: https://git.openjdk.java.net/valhalla/pull/623/files
  - new: https://git.openjdk.java.net/valhalla/pull/623/files/cb0dd95a..a3ccb572

Webrevs:
 - full: https://webrevs.openjdk.java.net/?repo=valhalla&pr=623&range=01
 - incr: https://webrevs.openjdk.java.net/?repo=valhalla&pr=623&range=00-01

  Stats: 25 lines in 6 files changed: 7 ins; 11 del; 7 mod
  Patch: https://git.openjdk.java.net/valhalla/pull/623.diff
  Fetch: git fetch https://git.openjdk.java.net/valhalla pull/623/head:pull/623

PR: https://git.openjdk.java.net/valhalla/pull/623

From fparain at openjdk.java.net  Wed Feb  2 15:58:23 2022
From: fparain at openjdk.java.net (Frederic Parain)
Date: Wed, 2 Feb 2022 15:58:23 GMT
Subject: [lworld] RFR: 8281116: [lworld] Adding Preload attribute support
 [v2]
In-Reply-To: <zlOl9xgtng_-Uhdkb_asJEK39E1SnTW-LfzBYos2e0s=.78e579c5-aba5-44a8-9634-6ccf2461c55e@github.com>
References: <CRTIZH_sHFnIXrvapvslAal1vhVNTsOdTNYBpJ-V7D0=.e2b4b5f2-1eae-4012-b266-e05e2133f477@github.com>
 <zlOl9xgtng_-Uhdkb_asJEK39E1SnTW-LfzBYos2e0s=.78e579c5-aba5-44a8-9634-6ccf2461c55e@github.com>
Message-ID: <QEouDtAUcbubwEIqfv85gtRQ0OqTfdEYbe9d3n1CPEQ=.c78e02b5-2b5c-4060-aff5-527002b2ae89@github.com>

On Wed, 2 Feb 2022 14:22:47 GMT, Harold Seigel <hseigel at openjdk.org> wrote:

>> Frederic Parain has updated the pull request incrementally with one additional commit since the last revision:
>> 
>>   Fixes to address Harold's comments
>
> src/hotspot/share/classfile/classFileParser.cpp line 5951:
> 
>> 5949:   _record_components(NULL),
>> 5950:   _temp_local_interfaces(NULL),
>> 5951:   _local_interfaces(NULL),
> 
> Do you need to free _preload_classes in ClassFileParser::~ClassFileParser() and set _preload_classes to NULL in ClassFileParser::clear_class_metadata(), also, assert that it is NULL in ClassFileParser::fill_instance_klass() ?

Fixed

> test/hotspot/jtreg/runtime/valhalla/inlinetypes/ValuePreloadTest.java line 61:
> 
>> 59:         checkFor(pb, "[warning][class,preload] Preloading of class PreloadValue1 during linking of class ValuePreloadClient1 (Preload attribute) failed");
>> 60:     }
>> 61: }
> 
> The test looks good but have multipe 2022's in the copyright:
>  * Copyright (c) 2022, 2022, Oracle and/or its affiliates. All rights reserved.

Fixed

-------------

PR: https://git.openjdk.java.net/valhalla/pull/623

From fparain at openjdk.java.net  Wed Feb  2 16:03:32 2022
From: fparain at openjdk.java.net (Frederic Parain)
Date: Wed, 2 Feb 2022 16:03:32 GMT
Subject: [lworld] RFR: 8281116: [lworld] Adding Preload attribute support
 [v2]
In-Reply-To: <IctDcuIUhv7s1dfaocc6sHdAatFYuoC7jrFKZ2JYNdw=.0a22c9aa-20da-46e3-8bb7-60ca10997f03@github.com>
References: <CRTIZH_sHFnIXrvapvslAal1vhVNTsOdTNYBpJ-V7D0=.e2b4b5f2-1eae-4012-b266-e05e2133f477@github.com>
 <IctDcuIUhv7s1dfaocc6sHdAatFYuoC7jrFKZ2JYNdw=.0a22c9aa-20da-46e3-8bb7-60ca10997f03@github.com>
Message-ID: <N9mkkybw_SK6Qdfq7vBjLQxvfw5PLSdSj_ojZvdQprM=.bab95a96-3ba1-4ffd-9429-dab80bfaea0f@github.com>

On Wed, 2 Feb 2022 15:58:22 GMT, Frederic Parain <fparain at openjdk.org> wrote:

>> Please review this changeset adding support for the Preload attribute.
>> It includes a logging feature to track and diagnose those optimistic class loading.
>> 
>> Tested with Mach5 tier1.
>> 
>> Thank you,
>> 
>> Fred
>
> Frederic Parain has updated the pull request incrementally with one additional commit since the last revision:
> 
>   Fixes to address Harold's comments

Harold,

Thank you for your feedback. Those issues should be fixed with the last commit.

Fred

-------------

PR: https://git.openjdk.java.net/valhalla/pull/623

From hseigel at openjdk.java.net  Wed Feb  2 16:09:36 2022
From: hseigel at openjdk.java.net (Harold Seigel)
Date: Wed, 2 Feb 2022 16:09:36 GMT
Subject: [lworld] RFR: 8281116: [lworld] Adding Preload attribute support
 [v2]
In-Reply-To: <IctDcuIUhv7s1dfaocc6sHdAatFYuoC7jrFKZ2JYNdw=.0a22c9aa-20da-46e3-8bb7-60ca10997f03@github.com>
References: <CRTIZH_sHFnIXrvapvslAal1vhVNTsOdTNYBpJ-V7D0=.e2b4b5f2-1eae-4012-b266-e05e2133f477@github.com>
 <IctDcuIUhv7s1dfaocc6sHdAatFYuoC7jrFKZ2JYNdw=.0a22c9aa-20da-46e3-8bb7-60ca10997f03@github.com>
Message-ID: <vSyCU74V-Wre1_IVDbnJqVbAGIi2HWjrdiNOjBv_FW4=.a5b28854-737c-4ea8-a837-045752232df1@github.com>

On Wed, 2 Feb 2022 15:58:22 GMT, Frederic Parain <fparain at openjdk.org> wrote:

>> Please review this changeset adding support for the Preload attribute.
>> It includes a logging feature to track and diagnose those optimistic class loading.
>> 
>> Tested with Mach5 tier1.
>> 
>> Thank you,
>> 
>> Fred
>
> Frederic Parain has updated the pull request incrementally with one additional commit since the last revision:
> 
>   Fixes to address Harold's comments

The changes look good!
Thanks, Harold

-------------

Marked as reviewed by hseigel (Committer).

PR: https://git.openjdk.java.net/valhalla/pull/623

From fparain at openjdk.java.net  Wed Feb  2 16:09:38 2022
From: fparain at openjdk.java.net (Frederic Parain)
Date: Wed, 2 Feb 2022 16:09:38 GMT
Subject: [lworld] Integrated: 8281116: [lworld] Adding Preload attribute
 support
In-Reply-To: <CRTIZH_sHFnIXrvapvslAal1vhVNTsOdTNYBpJ-V7D0=.e2b4b5f2-1eae-4012-b266-e05e2133f477@github.com>
References: <CRTIZH_sHFnIXrvapvslAal1vhVNTsOdTNYBpJ-V7D0=.e2b4b5f2-1eae-4012-b266-e05e2133f477@github.com>
Message-ID: <tH0G73ijv4vlSo7Qk-sbaG4DBnl_i34MXenpGmFzN7M=.5c0cf95e-6f13-48d9-873d-5b638c0abac5@github.com>

On Wed, 2 Feb 2022 13:12:19 GMT, Frederic Parain <fparain at openjdk.org> wrote:

> Please review this changeset adding support for the Preload attribute.
> It includes a logging feature to track and diagnose those optimistic class loading.
> 
> Tested with Mach5 tier1.
> 
> Thank you,
> 
> Fred

This pull request has now been integrated.

Changeset: f5a5dc35
Author:    Frederic Parain <fparain at openjdk.org>
URL:       https://git.openjdk.java.net/valhalla/commit/f5a5dc35f126734401eded910420ac5b805f0193
Stats:     371 lines in 10 files changed: 368 ins; 0 del; 3 mod

8281116: [lworld] Adding Preload attribute support

Reviewed-by: hseigel

-------------

PR: https://git.openjdk.java.net/valhalla/pull/623

From mchung at openjdk.java.net  Thu Feb  3 20:27:59 2022
From: mchung at openjdk.java.net (Mandy Chung)
Date: Thu, 3 Feb 2022 20:27:59 GMT
Subject: [lworld] RFR: 8280980: [lworld] MethodHandle and VarHandle support for
 value classes
Message-ID: <5F0j9coadl1nTba1hWGb_Y3yDWHSQhgXns_8qlLuAdg=.6a06004b-8e82-4749-ae12-40316fe552d1@github.com>

Add new test cases for value classes.

-------------

Commit messages:
 - Fix StaticFactoryTest not to run noInflation mode
 - 8280980: [lworld] MethodHandle and VarHandle support for value classes

Changes: https://git.openjdk.java.net/valhalla/pull/624/files
 Webrev: https://webrevs.openjdk.java.net/?repo=valhalla&pr=624&range=00
  Issue: https://bugs.openjdk.java.net/browse/JDK-8280980
  Stats: 432 lines in 10 files changed: 228 ins; 137 del; 67 mod
  Patch: https://git.openjdk.java.net/valhalla/pull/624.diff
  Fetch: git fetch https://git.openjdk.java.net/valhalla pull/624/head:pull/624

PR: https://git.openjdk.java.net/valhalla/pull/624

From mchung at openjdk.java.net  Thu Feb  3 20:34:40 2022
From: mchung at openjdk.java.net (Mandy Chung)
Date: Thu, 3 Feb 2022 20:34:40 GMT
Subject: [lworld] Integrated: 8280980: [lworld] MethodHandle and VarHandle
 support for value classes
In-Reply-To: <5F0j9coadl1nTba1hWGb_Y3yDWHSQhgXns_8qlLuAdg=.6a06004b-8e82-4749-ae12-40316fe552d1@github.com>
References: <5F0j9coadl1nTba1hWGb_Y3yDWHSQhgXns_8qlLuAdg=.6a06004b-8e82-4749-ae12-40316fe552d1@github.com>
Message-ID: <FDgTPGmE2lNH31Q12xhu9zer1sRJkDoFQTGu6LaiTbs=.2e6b8706-0e6f-4fdf-bdf6-0b3df0d6df14@github.com>

On Thu, 3 Feb 2022 20:21:54 GMT, Mandy Chung <mchung at openjdk.org> wrote:

> Add new test cases for value classes.

This pull request has now been integrated.

Changeset: b7a7fa72
Author:    Mandy Chung <mchung at openjdk.org>
URL:       https://git.openjdk.java.net/valhalla/commit/b7a7fa729f922b784e8167befe1ec786ec163bef
Stats:     432 lines in 10 files changed: 228 ins; 137 del; 67 mod

8280980: [lworld] MethodHandle and VarHandle support for value classes

-------------

PR: https://git.openjdk.java.net/valhalla/pull/624

From jespersm at openjdk.java.net  Fri Feb  4 09:50:52 2022
From: jespersm at openjdk.java.net (Jesper Steen =?UTF-8?B?TcO4bGxlcg==?=)
Date: Fri, 4 Feb 2022 09:50:52 GMT
Subject: [lworld] RFR: 8279840 [lworld] Inconsistent treatment of repeated
 modifiers
Message-ID: <OS9PWvr2joKJZRI45HZzDkEFEX5RGDLoW-7WdpDO1Mc=.d45d991e-ed59-4b08-a6cb-6a28b85c6448@github.com>

Make context-dependent keywords value and primitive act more like ordinary modifiers (for consistent error reporting.)

-------------

Commit messages:
 - 8279840 [lworld] Inconsistent treatment of repeated modifiers

Changes: https://git.openjdk.java.net/valhalla/pull/625/files
 Webrev: https://webrevs.openjdk.java.net/?repo=valhalla&pr=625&range=00
  Issue: https://bugs.openjdk.java.net/browse/JDK-8279840
  Stats: 33 lines in 3 files changed: 29 ins; 0 del; 4 mod
  Patch: https://git.openjdk.java.net/valhalla/pull/625.diff
  Fetch: git fetch https://git.openjdk.java.net/valhalla pull/625/head:pull/625

PR: https://git.openjdk.java.net/valhalla/pull/625

From fparain at openjdk.java.net  Fri Feb  4 17:06:07 2022
From: fparain at openjdk.java.net (Frederic Parain)
Date: Fri, 4 Feb 2022 17:06:07 GMT
Subject: [lworld] RFR: 8281281: [lworld] Preload attribute code needs some
 fixes
Message-ID: <uFMZQMJbyQR5D5hRztEm_ZiL09R5tPqBGMEyduotocA=.f30cab22-e205-4395-99ee-01b6f42a5439@github.com>

Please review those small fixes in the Preload attribute code.

Thank you,

Fred

-------------

Commit messages:
 - Various fixes in Preload attribute support

Changes: https://git.openjdk.java.net/valhalla/pull/626/files
 Webrev: https://webrevs.openjdk.java.net/?repo=valhalla&pr=626&range=00
  Issue: https://bugs.openjdk.java.net/browse/JDK-8281281
  Stats: 58 lines in 5 files changed: 50 ins; 4 del; 4 mod
  Patch: https://git.openjdk.java.net/valhalla/pull/626.diff
  Fetch: git fetch https://git.openjdk.java.net/valhalla pull/626/head:pull/626

PR: https://git.openjdk.java.net/valhalla/pull/626

From hseigel at openjdk.java.net  Fri Feb  4 18:51:34 2022
From: hseigel at openjdk.java.net (Harold Seigel)
Date: Fri, 4 Feb 2022 18:51:34 GMT
Subject: [lworld] RFR: 8281281: [lworld] Preload attribute code needs some
 fixes
In-Reply-To: <uFMZQMJbyQR5D5hRztEm_ZiL09R5tPqBGMEyduotocA=.f30cab22-e205-4395-99ee-01b6f42a5439@github.com>
References: <uFMZQMJbyQR5D5hRztEm_ZiL09R5tPqBGMEyduotocA=.f30cab22-e205-4395-99ee-01b6f42a5439@github.com>
Message-ID: <hN37dGr0bFzL1NZJlT6GrmIvuodDstYVcj4nI5Stfvk=.fed5e573-b679-45a5-b758-c8a53bbe5e9c@github.com>

On Fri, 4 Feb 2022 16:58:22 GMT, Frederic Parain <fparain at openjdk.org> wrote:

> Please review those small fixes in the Preload attribute code.
> 
> Thank you,
> 
> Fred

Changes look good!
Thanks, Harold

-------------

Marked as reviewed by hseigel (Committer).

PR: https://git.openjdk.java.net/valhalla/pull/626

From fparain at openjdk.java.net  Fri Feb  4 19:11:28 2022
From: fparain at openjdk.java.net (Frederic Parain)
Date: Fri, 4 Feb 2022 19:11:28 GMT
Subject: git: openjdk/valhalla: lworld: 8281281: [lworld] Preload attribute
 code needs some fixes
Message-ID: <2e81a7e8-3837-4252-9cc0-24fc1ce38b5a@openjdk.org>

Changeset: a7b10d2a
Author:    Frederic Parain <fparain at openjdk.org>
Date:      2022-02-04 19:11:05 +0000
URL:       https://git.openjdk.java.net/valhalla/commit/a7b10d2acff754b015a4372b4b838d1d41b39940

8281281: [lworld] Preload attribute code needs some fixes

Reviewed-by: hseigel

! src/hotspot/share/classfile/classFileParser.cpp
! src/hotspot/share/oops/instanceKlass.cpp
! src/hotspot/share/oops/instanceKlass.hpp
! src/hotspot/share/prims/jvmtiRedefineClasses.cpp
! src/hotspot/share/prims/jvmtiRedefineClasses.hpp


From fparain at openjdk.java.net  Fri Feb  4 19:13:39 2022
From: fparain at openjdk.java.net (Frederic Parain)
Date: Fri, 4 Feb 2022 19:13:39 GMT
Subject: [lworld] RFR: 8281281: [lworld] Preload attribute code needs some
 fixes
In-Reply-To: <uFMZQMJbyQR5D5hRztEm_ZiL09R5tPqBGMEyduotocA=.f30cab22-e205-4395-99ee-01b6f42a5439@github.com>
References: <uFMZQMJbyQR5D5hRztEm_ZiL09R5tPqBGMEyduotocA=.f30cab22-e205-4395-99ee-01b6f42a5439@github.com>
Message-ID: <lK1LMmWmsKcpYNtaBtfP3fXY8EC7tS1WLYQbohpl0H8=.225707f7-99c5-4ca4-85ef-1e6ff8982ecb@github.com>

On Fri, 4 Feb 2022 16:58:22 GMT, Frederic Parain <fparain at openjdk.org> wrote:

> Please review those small fixes in the Preload attribute code.
> 
> Thank you,
> 
> Fred

Thank you for the review Harold.

-------------

PR: https://git.openjdk.java.net/valhalla/pull/626

From fparain at openjdk.java.net  Fri Feb  4 19:13:40 2022
From: fparain at openjdk.java.net (Frederic Parain)
Date: Fri, 4 Feb 2022 19:13:40 GMT
Subject: [lworld] Integrated: 8281281: [lworld] Preload attribute code needs
 some fixes
In-Reply-To: <uFMZQMJbyQR5D5hRztEm_ZiL09R5tPqBGMEyduotocA=.f30cab22-e205-4395-99ee-01b6f42a5439@github.com>
References: <uFMZQMJbyQR5D5hRztEm_ZiL09R5tPqBGMEyduotocA=.f30cab22-e205-4395-99ee-01b6f42a5439@github.com>
Message-ID: <oTqqDhde5Pg_LsKpftWQuFnjNOzzionWRVkDzyJiHo8=.edfa3434-78ff-48ea-a99e-70a833bf3510@github.com>

On Fri, 4 Feb 2022 16:58:22 GMT, Frederic Parain <fparain at openjdk.org> wrote:

> Please review those small fixes in the Preload attribute code.
> 
> Thank you,
> 
> Fred

This pull request has now been integrated.

Changeset: a7b10d2a
Author:    Frederic Parain <fparain at openjdk.org>
URL:       https://git.openjdk.java.net/valhalla/commit/a7b10d2acff754b015a4372b4b838d1d41b39940
Stats:     58 lines in 5 files changed: 50 ins; 4 del; 4 mod

8281281: [lworld] Preload attribute code needs some fixes

Reviewed-by: hseigel

-------------

PR: https://git.openjdk.java.net/valhalla/pull/626

From vromero at openjdk.java.net  Fri Feb  4 22:21:06 2022
From: vromero at openjdk.java.net (Vicente Romero)
Date: Fri, 4 Feb 2022 22:21:06 GMT
Subject: Integrated: Adding type system regression tests
Message-ID: <UZtrdoO-Dbf5814nnL-0pDZ0JxnEb_AAs69aQI974_A=.0430db7b-fffa-41fb-8e74-fb40a153c8da@github.com>

Adding more type system regression tests for universal type variables. Adding overload resolution for primitive classes.

-------------

Commit messages:
 - inference related tests
 - adding type system regression tests

Changes: https://git.openjdk.java.net/valhalla/pull/628/files
 Webrev: https://webrevs.openjdk.java.net/?repo=valhalla&pr=628&range=00
  Stats: 297 lines in 10 files changed: 223 ins; 26 del; 48 mod
  Patch: https://git.openjdk.java.net/valhalla/pull/628.diff
  Fetch: git fetch https://git.openjdk.java.net/valhalla pull/628/head:pull/628

PR: https://git.openjdk.java.net/valhalla/pull/628

From vromero at openjdk.java.net  Fri Feb  4 22:21:06 2022
From: vromero at openjdk.java.net (Vicente Romero)
Date: Fri, 4 Feb 2022 22:21:06 GMT
Subject: Integrated: Adding type system regression tests
In-Reply-To: <UZtrdoO-Dbf5814nnL-0pDZ0JxnEb_AAs69aQI974_A=.0430db7b-fffa-41fb-8e74-fb40a153c8da@github.com>
References: <UZtrdoO-Dbf5814nnL-0pDZ0JxnEb_AAs69aQI974_A=.0430db7b-fffa-41fb-8e74-fb40a153c8da@github.com>
Message-ID: <IbS5SUgdKhw_g3gAvQY5608qiw0Pf2gwaFRgBIRdIts=.4b5291f5-9cc9-4cbd-a4d1-a81e7c6ca93f@github.com>

On Fri, 4 Feb 2022 22:11:23 GMT, Vicente Romero <vromero at openjdk.org> wrote:

> Adding more type system regression tests for universal type variables. Adding overload resolution for primitive classes.

This pull request has now been integrated.

Changeset: cf21b754
Author:    Vicente Romero <vromero at openjdk.org>
URL:       https://git.openjdk.java.net/valhalla/commit/cf21b7542c9f5cff54c07a526cfa5764d42169e8
Stats:     297 lines in 10 files changed: 223 ins; 26 del; 48 mod

Adding type system regression tests

-------------

PR: https://git.openjdk.java.net/valhalla/pull/628

From jespersm at openjdk.java.net  Fri Feb  4 23:24:59 2022
From: jespersm at openjdk.java.net (Jesper Steen =?UTF-8?B?TcO4bGxlcg==?=)
Date: Fri, 4 Feb 2022 23:24:59 GMT
Subject: [lworld] RFR: 8279839 [lworld] Javac has started incorrectly accepting
 native as a modifer for classes
Message-ID: <LSRBDbntKIsICYhAtsJXUjHGETAkghY870XivERUF08=.539819c8-2891-4717-8442-c2acb7948f51@github.com>

Confusion between the ACC_VALUE and ACC_NATIVE resolved
by making explicit class flags masks for the adjusted flags.

-------------

Commit messages:
 - Stray whitespace
 - 8279839 [lworld] Javac has started incorrectly accepting native as a modifer for classes

Changes: https://git.openjdk.java.net/valhalla/pull/627/files
 Webrev: https://webrevs.openjdk.java.net/?repo=valhalla&pr=627&range=00
  Issue: https://bugs.openjdk.java.net/browse/JDK-8279839
  Stats: 21 lines in 5 files changed: 17 ins; 0 del; 4 mod
  Patch: https://git.openjdk.java.net/valhalla/pull/627.diff
  Fetch: git fetch https://git.openjdk.java.net/valhalla pull/627/head:pull/627

PR: https://git.openjdk.java.net/valhalla/pull/627

From jespersm at openjdk.java.net  Sat Feb  5 11:48:35 2022
From: jespersm at openjdk.java.net (Jesper Steen =?UTF-8?B?TcO4bGxlcg==?=)
Date: Sat, 5 Feb 2022 11:48:35 GMT
Subject: [lworld] RFR: 8279839 [lworld] Javac has started incorrectly
 accepting native as a modifer for classes
In-Reply-To: <LSRBDbntKIsICYhAtsJXUjHGETAkghY870XivERUF08=.539819c8-2891-4717-8442-c2acb7948f51@github.com>
References: <LSRBDbntKIsICYhAtsJXUjHGETAkghY870XivERUF08=.539819c8-2891-4717-8442-c2acb7948f51@github.com>
Message-ID: <tvk7pZ1lhLzvWjKYdakCSXo2tCwPpIEtW0fI-zk2z-M=.ec33c7bd-aacc-487e-a3c0-54323ab919d7@github.com>

On Fri, 4 Feb 2022 21:42:33 GMT, Jesper Steen M?ller <jespersm at openjdk.org> wrote:

> Confusion between the ACC_VALUE and ACC_NATIVE resolved
> by making explicit class flags masks for the adjusted flags.

To the reviewer: This time, I tried the "integrate auto" command since it's a small change and this would smooth out the stuttering "approve - integrate - sponsor" cycle across timezones.
I don't know it that is how it is intended, and a bit surpriset it kicked off another round of checks.

-------------

PR: https://git.openjdk.java.net/valhalla/pull/627

From sadayapalam at openjdk.java.net  Mon Feb  7 07:41:06 2022
From: sadayapalam at openjdk.java.net (Srikanth Adayapalam)
Date: Mon, 7 Feb 2022 07:41:06 GMT
Subject: [lworld] RFR: 8281166: [lworld] javac should generate BSM to
 invoke the static factory for value class
In-Reply-To: <1fGVM7EAm5XXVEuslIbog5A_my4p6s61UnCWngABxpU=.8550588d-cd92-45d9-ac61-67c44772827f@github.com>
References: <1fGVM7EAm5XXVEuslIbog5A_my4p6s61UnCWngABxpU=.8550588d-cd92-45d9-ac61-67c44772827f@github.com>
Message-ID: <Oe5oHPIIE9jgnqvaKXLif2nR5QUFFrcLn_X4IqBELrE=.c7ca7414-bfaf-4303-93e6-eb436931c436@github.com>

On Mon, 7 Feb 2022 07:24:36 GMT, Srikanth Adayapalam <sadayapalam at openjdk.org> wrote:

> When constructor reference expressions invole value classes, ensure that TransValues will see the constructor call in order for that call to be transformed into factory method invocation.

src/jdk.compiler/share/classes/com/sun/tools/javac/comp/LambdaToMethod.java line 2345:

> 2343:                           tree.kind != ReferenceKind.ARRAY_CTOR &&
> 2344:                           (tree.sym.owner.isDirectlyOrIndirectlyLocal() || tree.sym.owner.isInner() || tree.sym.owner.isValueClass()));
> 2345:             }

Notes for reviewer: Basically by forcing the constructor references to be converted into lambdas, we ensure that later pipeline stages will see the constructor invocation. For example, with the code change above, given a value class V, a constructor reference of the form V::new will be transformed into a lambda implementation method of the form 


/*synthetic*/ private static X$V lambda$main$1() {
    return new X$V();
}

This lambda implementation method has the constructor invocation intact which can then be detected and transformed by *com.sun.tools.javac.jvm.TransValues#visitNewClass*

(It is possible to directly emit BSM entries without converting them to lambda but that would result in code duplication - we would have to look at DynamicMethod symbols, pick them apart and reencode static and dynamic arguments to indy so that value constructors (and primitive class constructors are reencoded to static factory)

-------------

PR: https://git.openjdk.java.net/valhalla/pull/629

From sadayapalam at openjdk.java.net  Mon Feb  7 07:41:05 2022
From: sadayapalam at openjdk.java.net (Srikanth Adayapalam)
Date: Mon, 7 Feb 2022 07:41:05 GMT
Subject: [lworld] RFR: 8281166: [lworld] javac should generate BSM to invoke
 the static factory for value class
Message-ID: <1fGVM7EAm5XXVEuslIbog5A_my4p6s61UnCWngABxpU=.8550588d-cd92-45d9-ac61-67c44772827f@github.com>

When constructor reference expressions invole value classes, ensure that TransValues will see the constructor call in order for that call to be transformed into factory method invocation.

-------------

Commit messages:
 - Fix white space errors
 - 8281166: javac should generate BSM to invoke the static factory for value class

Changes: https://git.openjdk.java.net/valhalla/pull/629/files
 Webrev: https://webrevs.openjdk.java.net/?repo=valhalla&pr=629&range=00
  Issue: https://bugs.openjdk.java.net/browse/JDK-8281166
  Stats: 72 lines in 2 files changed: 71 ins; 0 del; 1 mod
  Patch: https://git.openjdk.java.net/valhalla/pull/629.diff
  Fetch: git fetch https://git.openjdk.java.net/valhalla pull/629/head:pull/629

PR: https://git.openjdk.java.net/valhalla/pull/629

From forax at openjdk.java.net  Mon Feb  7 08:00:38 2022
From: forax at openjdk.java.net (=?UTF-8?B?UsOpbWk=?= Forax)
Date: Mon, 7 Feb 2022 08:00:38 GMT
Subject: [lworld] RFR: 8281166: [lworld] javac should generate BSM to
 invoke the static factory for value class
In-Reply-To: <Oe5oHPIIE9jgnqvaKXLif2nR5QUFFrcLn_X4IqBELrE=.c7ca7414-bfaf-4303-93e6-eb436931c436@github.com>
References: <1fGVM7EAm5XXVEuslIbog5A_my4p6s61UnCWngABxpU=.8550588d-cd92-45d9-ac61-67c44772827f@github.com>
 <Oe5oHPIIE9jgnqvaKXLif2nR5QUFFrcLn_X4IqBELrE=.c7ca7414-bfaf-4303-93e6-eb436931c436@github.com>
Message-ID: <utk75UUvRlgpUTnggOPjnYRwlmXN-EuuhpILte84w0A=.c19fa40a-1c6d-4a0c-ae5b-19f0a6479196@github.com>

On Mon, 7 Feb 2022 07:32:28 GMT, Srikanth Adayapalam <sadayapalam at openjdk.org> wrote:

>> When constructor reference expressions invole value classes, ensure that TransValues will see the constructor call in order for that call to be transformed into factory method invocation.
>
> src/jdk.compiler/share/classes/com/sun/tools/javac/comp/LambdaToMethod.java line 2345:
> 
>> 2343:                           tree.kind != ReferenceKind.ARRAY_CTOR &&
>> 2344:                           (tree.sym.owner.isDirectlyOrIndirectlyLocal() || tree.sym.owner.isInner() || tree.sym.owner.isValueClass()));
>> 2345:             }
> 
> Notes for reviewer: Basically by forcing the constructor references to be converted into lambdas, we ensure that later pipeline stages will see the constructor invocation. For example, with the code change above, given a value class V, a constructor reference of the form V::new will be transformed into a lambda implementation method of the form 
> 
> 
> /*synthetic*/ private static X$V lambda$main$1() {
>     return new X$V();
> }
> 
> This lambda implementation method has the constructor invocation intact which can then be detected and transformed by *com.sun.tools.javac.jvm.TransValues#visitNewClass*
> 
> (It is possible to directly emit BSM entries without converting them to lambda but that would result in code duplication - we would have to look at DynamicMethod symbols, pick them apart and reencode static and dynamic arguments to indy so that value constructors (and primitive class constructors are reencoded to static factory)

I wonder if the transformation from a constructor call to a static method call when it's a value class should not be done before the lambda transformation ?

-------------

PR: https://git.openjdk.java.net/valhalla/pull/629

From sadayapalam at openjdk.java.net  Mon Feb  7 08:08:37 2022
From: sadayapalam at openjdk.java.net (Srikanth Adayapalam)
Date: Mon, 7 Feb 2022 08:08:37 GMT
Subject: [lworld] RFR: 8279839 [lworld] Javac has started incorrectly
 accepting native as a modifer for classes
In-Reply-To: <LSRBDbntKIsICYhAtsJXUjHGETAkghY870XivERUF08=.539819c8-2891-4717-8442-c2acb7948f51@github.com>
References: <LSRBDbntKIsICYhAtsJXUjHGETAkghY870XivERUF08=.539819c8-2891-4717-8442-c2acb7948f51@github.com>
Message-ID: <1nIOfqSMdVDfaH-IeevrOT6nF5Mgs5WYfjSKLBlr1Xc=.6deef43b-e82c-469c-ada8-6b3793677bbd@github.com>

On Fri, 4 Feb 2022 21:42:33 GMT, Jesper Steen M?ller <jespersm at openjdk.org> wrote:

> Confusion between the ACC_VALUE and ACC_NATIVE resolved
> by making explicit class flags masks for the adjusted flags.

Changes look good - Thanks Jesper!

-------------

Marked as reviewed by sadayapalam (Committer).

PR: https://git.openjdk.java.net/valhalla/pull/627

From jespersm at openjdk.java.net  Mon Feb  7 08:12:38 2022
From: jespersm at openjdk.java.net (Jesper Steen =?UTF-8?B?TcO4bGxlcg==?=)
Date: Mon, 7 Feb 2022 08:12:38 GMT
Subject: [lworld] Integrated: 8279839 [lworld] Javac has started incorrectly
 accepting native as a modifer for classes
In-Reply-To: <LSRBDbntKIsICYhAtsJXUjHGETAkghY870XivERUF08=.539819c8-2891-4717-8442-c2acb7948f51@github.com>
References: <LSRBDbntKIsICYhAtsJXUjHGETAkghY870XivERUF08=.539819c8-2891-4717-8442-c2acb7948f51@github.com>
Message-ID: <YpRCVM83m684DG1cHqVI1gy6EqtmjFBeCZLUDPInOJg=.f55ab6e1-01a4-4844-a2b6-195522a0303d@github.com>

On Fri, 4 Feb 2022 21:42:33 GMT, Jesper Steen M?ller <jespersm at openjdk.org> wrote:

> Confusion between the ACC_VALUE and ACC_NATIVE resolved
> by making explicit class flags masks for the adjusted flags.

This pull request has now been integrated.

Changeset: 6e609b32
Author:    Jesper Steen M?ller <jespersm at openjdk.org>
Committer: Srikanth Adayapalam <sadayapalam at openjdk.org>
URL:       https://git.openjdk.java.net/valhalla/commit/6e609b324aed08413c9eae2e7cc93814c8d24933
Stats:     21 lines in 5 files changed: 17 ins; 0 del; 4 mod

8279839: [lworld] Javac has started incorrectly accepting native as a modifer for classes

Reviewed-by: sadayapalam

-------------

PR: https://git.openjdk.java.net/valhalla/pull/627

From sadayapalam at openjdk.java.net  Mon Feb  7 08:15:34 2022
From: sadayapalam at openjdk.java.net (Srikanth Adayapalam)
Date: Mon, 7 Feb 2022 08:15:34 GMT
Subject: [lworld] RFR: 8281166: [lworld] javac should generate BSM to
 invoke the static factory for value class
In-Reply-To: <utk75UUvRlgpUTnggOPjnYRwlmXN-EuuhpILte84w0A=.c19fa40a-1c6d-4a0c-ae5b-19f0a6479196@github.com>
References: <1fGVM7EAm5XXVEuslIbog5A_my4p6s61UnCWngABxpU=.8550588d-cd92-45d9-ac61-67c44772827f@github.com>
 <Oe5oHPIIE9jgnqvaKXLif2nR5QUFFrcLn_X4IqBELrE=.c7ca7414-bfaf-4303-93e6-eb436931c436@github.com>
 <utk75UUvRlgpUTnggOPjnYRwlmXN-EuuhpILte84w0A=.c19fa40a-1c6d-4a0c-ae5b-19f0a6479196@github.com>
Message-ID: <fJcaKsq4neP0_naWfjaNuKcJMb--D2CngTQVwTxXcx0=.b2aa42f0-b478-4c5e-b8e8-956c1337010b@github.com>

On Mon, 7 Feb 2022 07:57:43 GMT, R?mi Forax <forax at openjdk.org> wrote:

> I wonder if the transformation from a constructor call to a static method call when it's a value class should not be done before the lambda transformation ?

This is technically feasible, but has some practical complications. As the top level comment in com.sun.tools.javac.jvm.TransValues calls out:


/**
 * This pass translates value class constructors into static factory methods and patches up constructor
 * calls to become invocations of those static factory methods.
 *
 * We get commissioned as a subpass of Gen. Constructor trees undergo plenty of change in Lower
 * (enclosing instance injection, captured locals ...) and in Gen (instance field initialization,
 * see normalizeDefs) and so it is most effective to wait until things reach a quiescent state
 * before undertaking the tinkering that we do.
 *
 * See https://bugs.openjdk.java.net/browse/JDK-8198749 for the kind of transformations we do.
 *
 */


Lambda lowering happens well before Lower and Gen pipeline phases. TransValues is a subphase of Gen - we need to wait until Gen because constructors undergo lots of changes up until and well into Gen. So if we want to group all transformations - of constructors into static factories and constructor invocations into static value factory invocations - into one centralized localized pipeline stage (i.e TransValues) the present approach is the best

-------------

PR: https://git.openjdk.java.net/valhalla/pull/629

From sadayapalam at openjdk.java.net  Mon Feb  7 08:44:36 2022
From: sadayapalam at openjdk.java.net (Srikanth Adayapalam)
Date: Mon, 7 Feb 2022 08:44:36 GMT
Subject: [lworld] RFR: 8279840 [lworld] Inconsistent treatment of repeated
 modifiers.
In-Reply-To: <OS9PWvr2joKJZRI45HZzDkEFEX5RGDLoW-7WdpDO1Mc=.d45d991e-ed59-4b08-a6cb-6a28b85c6448@github.com>
References: <OS9PWvr2joKJZRI45HZzDkEFEX5RGDLoW-7WdpDO1Mc=.d45d991e-ed59-4b08-a6cb-6a28b85c6448@github.com>
Message-ID: <YxhR9neyOFDNoNi4gfwcR3YLH2FEzYRK8FcynBuz-oQ=.9914557c-f534-4d7a-beba-7500c42ac191@github.com>

On Fri, 4 Feb 2022 09:45:15 GMT, Jesper Steen M?ller <jespersm at openjdk.org> wrote:

> Make context-dependent keywords value and primitive act more like ordinary modifiers (for consistent error reporting.)

Changes look good - Thanks Jepser! Please proceed to integrate

-------------

Marked as reviewed by sadayapalam (Committer).

PR: https://git.openjdk.java.net/valhalla/pull/625

From dsimms at openjdk.java.net  Mon Feb  7 09:13:07 2022
From: dsimms at openjdk.java.net (David Simms)
Date: Mon, 7 Feb 2022 09:13:07 GMT
Subject: [lworld] RFR: Merge jdk
Message-ID: <OOwz1-2cCAZtkkQUsnyFlsBTNm-ObFmqrIzsmJWFdKA=.3a41a47d-4a5f-45ef-a896-bf7bcc4d3343@github.com>

Merge jdk-19+8

Further fixes for:
- CP indices modified by changes from mainline (javac tests)
- Finalizer deprecation
- Classfile version 63

-------------

Commit messages:
 - Further CP indices modified by changes from mainline
 - Finalizer deprecation
 - CP indices modified by changes from mainline
 - Version 63
 - Logical merge fixes
 - Merge tag 'jdk-19+8' into lworld_merge_jdk_19_8
 - 8278254: Cleanup doclint warnings in java.desktop module
 - 8280600: C2: assert(!had_error) failed: bad dominance
 - 8279535: C2: Dead code in PhaseIdealLoop::create_loop_nest after JDK-8276116
 - 8280885: Shenandoah: Some tests failed with "EA: missing allocation reference path"
 - ... and 855 more: https://git.openjdk.java.net/valhalla/compare/f5a5dc35...59d74953

The webrevs contain the adjustments done while merging with regards to each parent branch:
 - lworld: https://webrevs.openjdk.java.net/?repo=valhalla&pr=630&range=00.0
 - jdk: https://webrevs.openjdk.java.net/?repo=valhalla&pr=630&range=00.1

Changes: https://git.openjdk.java.net/valhalla/pull/630/files
  Stats: 103971 lines in 2818 files changed: 68416 ins; 20442 del; 15113 mod
  Patch: https://git.openjdk.java.net/valhalla/pull/630.diff
  Fetch: git fetch https://git.openjdk.java.net/valhalla pull/630/head:pull/630

PR: https://git.openjdk.java.net/valhalla/pull/630

From forax at openjdk.java.net  Mon Feb  7 09:32:28 2022
From: forax at openjdk.java.net (=?UTF-8?B?UsOpbWk=?= Forax)
Date: Mon, 7 Feb 2022 09:32:28 GMT
Subject: [lworld] RFR: 8281166: [lworld] javac should generate BSM to
 invoke the static factory for value class
In-Reply-To: <fJcaKsq4neP0_naWfjaNuKcJMb--D2CngTQVwTxXcx0=.b2aa42f0-b478-4c5e-b8e8-956c1337010b@github.com>
References: <1fGVM7EAm5XXVEuslIbog5A_my4p6s61UnCWngABxpU=.8550588d-cd92-45d9-ac61-67c44772827f@github.com>
 <Oe5oHPIIE9jgnqvaKXLif2nR5QUFFrcLn_X4IqBELrE=.c7ca7414-bfaf-4303-93e6-eb436931c436@github.com>
 <utk75UUvRlgpUTnggOPjnYRwlmXN-EuuhpILte84w0A=.c19fa40a-1c6d-4a0c-ae5b-19f0a6479196@github.com>
 <fJcaKsq4neP0_naWfjaNuKcJMb--D2CngTQVwTxXcx0=.b2aa42f0-b478-4c5e-b8e8-956c1337010b@github.com>
Message-ID: <rUw4ljrxBYTgohKoBAQUeG0mxcWA8bFM-Tq9Fppnu6w=.d820f96b-6e59-44b5-b2ba-ceb3d20aa45a@github.com>

On Mon, 7 Feb 2022 08:11:57 GMT, Srikanth Adayapalam <sadayapalam at openjdk.org> wrote:

>> I wonder if the transformation from a constructor call to a static method call when it's a value class should not be done before the lambda transformation ?
>
>> I wonder if the transformation from a constructor call to a static method call when it's a value class should not be done before the lambda transformation ?
> 
> This is technically feasible, but has some practical complications. As the top level comment in com.sun.tools.javac.jvm.TransValues calls out:
> 
> 
> /**
>  * This pass translates value class constructors into static factory methods and patches up constructor
>  * calls to become invocations of those static factory methods.
>  *
>  * We get commissioned as a subpass of Gen. Constructor trees undergo plenty of change in Lower
>  * (enclosing instance injection, captured locals ...) and in Gen (instance field initialization,
>  * see normalizeDefs) and so it is most effective to wait until things reach a quiescent state
>  * before undertaking the tinkering that we do.
>  *
>  * See https://bugs.openjdk.java.net/browse/JDK-8198749 for the kind of transformations we do.
>  *
>  */
> 
> 
> Lambda lowering happens well before Lower and Gen pipeline phases. TransValues is a subphase of Gen - we need to wait until Gen because constructors undergo lots of changes up until and well into Gen. So if we want to group all transformations - of constructors into static factories and constructor invocations into static value factory invocations - into one centralized localized pipeline stage (i.e TransValues) the present approach is the best

Okay, make sense

-------------

PR: https://git.openjdk.java.net/valhalla/pull/629

From dsimms at openjdk.java.net  Mon Feb  7 09:44:41 2022
From: dsimms at openjdk.java.net (David Simms)
Date: Mon, 7 Feb 2022 09:44:41 GMT
Subject: [lworld] RFR: Merge jdk [v2]
In-Reply-To: <OOwz1-2cCAZtkkQUsnyFlsBTNm-ObFmqrIzsmJWFdKA=.3a41a47d-4a5f-45ef-a896-bf7bcc4d3343@github.com>
References: <OOwz1-2cCAZtkkQUsnyFlsBTNm-ObFmqrIzsmJWFdKA=.3a41a47d-4a5f-45ef-a896-bf7bcc4d3343@github.com>
Message-ID: <DflLvgqXIGCvqJUu1ZGyl3A5ToNs7-4TC08Eyqw569Y=.bf3e3e33-a11b-41f4-9c9c-13771de559b4@github.com>

> Merge jdk-19+8
> 
> Further fixes for:
> - CP indices modified by changes from mainline (javac tests)
> - Finalizer deprecation
> - Classfile version 63

David Simms has updated the pull request incrementally with one additional commit since the last revision:

  JDK-8281325 TestCallingConvention problem listed

-------------

Changes:
  - all: https://git.openjdk.java.net/valhalla/pull/630/files
  - new: https://git.openjdk.java.net/valhalla/pull/630/files/59d74953..98c48116

Webrevs:
 - full: https://webrevs.openjdk.java.net/?repo=valhalla&pr=630&range=01
 - incr: https://webrevs.openjdk.java.net/?repo=valhalla&pr=630&range=00-01

  Stats: 2 lines in 1 file changed: 2 ins; 0 del; 0 mod
  Patch: https://git.openjdk.java.net/valhalla/pull/630.diff
  Fetch: git fetch https://git.openjdk.java.net/valhalla pull/630/head:pull/630

PR: https://git.openjdk.java.net/valhalla/pull/630

From dsimms at openjdk.java.net  Mon Feb  7 09:47:44 2022
From: dsimms at openjdk.java.net (David Simms)
Date: Mon, 7 Feb 2022 09:47:44 GMT
Subject: [lworld] Integrated: Merge jdk
In-Reply-To: <OOwz1-2cCAZtkkQUsnyFlsBTNm-ObFmqrIzsmJWFdKA=.3a41a47d-4a5f-45ef-a896-bf7bcc4d3343@github.com>
References: <OOwz1-2cCAZtkkQUsnyFlsBTNm-ObFmqrIzsmJWFdKA=.3a41a47d-4a5f-45ef-a896-bf7bcc4d3343@github.com>
Message-ID: <K5JNpkRB7tqjV_ojAaVoO3gNiylSxIWxGi5jjin6JI0=.038da0cc-7e05-4a4b-8655-b0ede7f45577@github.com>

On Mon, 7 Feb 2022 09:02:59 GMT, David Simms <dsimms at openjdk.org> wrote:

> Merge jdk-19+8
> 
> Further fixes for:
> - CP indices modified by changes from mainline (javac tests)
> - Finalizer deprecation
> - Classfile version 63

This pull request has now been integrated.

Changeset: 68b7cfaa
Author:    David Simms <dsimms at openjdk.org>
URL:       https://git.openjdk.java.net/valhalla/commit/68b7cfaa27339454f48571154da3545c9433e13e
Stats:     103973 lines in 2818 files changed: 68418 ins; 20442 del; 15113 mod

Merge jdk

Merge jdk-19+8

-------------

PR: https://git.openjdk.java.net/valhalla/pull/630

From duke at openjdk.java.net  Mon Feb  7 10:17:35 2022
From: duke at openjdk.java.net (Aggelos Biboudis)
Date: Mon, 7 Feb 2022 10:17:35 GMT
Subject: [lworld] RFR: 8281166: [lworld] javac should generate BSM to
 invoke the static factory for value class
In-Reply-To: <1fGVM7EAm5XXVEuslIbog5A_my4p6s61UnCWngABxpU=.8550588d-cd92-45d9-ac61-67c44772827f@github.com>
References: <1fGVM7EAm5XXVEuslIbog5A_my4p6s61UnCWngABxpU=.8550588d-cd92-45d9-ac61-67c44772827f@github.com>
Message-ID: <nPJOKWzVXBFW39dvwo-NlQ5dBT-FbIaKhODRBQ05RUA=.91031cb4-f31d-4f9e-ad73-8de31db785b2@github.com>

On Mon, 7 Feb 2022 07:24:36 GMT, Srikanth Adayapalam <sadayapalam at openjdk.org> wrote:

> When constructor reference expressions invole value classes, ensure that TransValues will see the constructor call in order for that call to be transformed into factory method invocation.

????

-------------

Marked as reviewed by biboudis at github.com (no known OpenJDK username).

PR: https://git.openjdk.java.net/valhalla/pull/629

From jespersm at openjdk.java.net  Mon Feb  7 10:47:23 2022
From: jespersm at openjdk.java.net (Jesper Steen =?UTF-8?B?TcO4bGxlcg==?=)
Date: Mon, 7 Feb 2022 10:47:23 GMT
Subject: [lworld] Integrated: 8279840 [lworld] Inconsistent treatment of
 repeated modifiers.
In-Reply-To: <OS9PWvr2joKJZRI45HZzDkEFEX5RGDLoW-7WdpDO1Mc=.d45d991e-ed59-4b08-a6cb-6a28b85c6448@github.com>
References: <OS9PWvr2joKJZRI45HZzDkEFEX5RGDLoW-7WdpDO1Mc=.d45d991e-ed59-4b08-a6cb-6a28b85c6448@github.com>
Message-ID: <iXRcRz3xBuOTSPBSyp7_ETS-4hweYB9bD5eWu25EXzc=.fca06fad-6b65-44b4-9f89-cb6cde5e956b@github.com>

On Fri, 4 Feb 2022 09:45:15 GMT, Jesper Steen M?ller <jespersm at openjdk.org> wrote:

> Make context-dependent keywords value and primitive act more like ordinary modifiers (for consistent error reporting.)

This pull request has now been integrated.

Changeset: 4ae0a0d2
Author:    Jesper Steen M?ller <jespersm at openjdk.org>
Committer: Srikanth Adayapalam <sadayapalam at openjdk.org>
URL:       https://git.openjdk.java.net/valhalla/commit/4ae0a0d2e11d67f6d1efe161674110aff171e872
Stats:     33 lines in 3 files changed: 29 ins; 0 del; 4 mod

8279840: [lworld] Inconsistent treatment of repeated modifiers.

Reviewed-by: sadayapalam

-------------

PR: https://git.openjdk.java.net/valhalla/pull/625

From sadayapalam at openjdk.java.net  Mon Feb  7 10:22:39 2022
From: sadayapalam at openjdk.java.net (Srikanth Adayapalam)
Date: Mon, 7 Feb 2022 10:22:39 GMT
Subject: [lworld] Integrated: 8281166: [lworld] javac should generate BSM to
 invoke the static factory for value class
In-Reply-To: <1fGVM7EAm5XXVEuslIbog5A_my4p6s61UnCWngABxpU=.8550588d-cd92-45d9-ac61-67c44772827f@github.com>
References: <1fGVM7EAm5XXVEuslIbog5A_my4p6s61UnCWngABxpU=.8550588d-cd92-45d9-ac61-67c44772827f@github.com>
Message-ID: <GsBkdSpQ3RtWh_bX1_h4M6DyV2-FGQSGRxbl-9TiK28=.3fb1f6f4-f308-4e2d-a465-7481d9d6084a@github.com>

On Mon, 7 Feb 2022 07:24:36 GMT, Srikanth Adayapalam <sadayapalam at openjdk.org> wrote:

> When constructor reference expressions invole value classes, ensure that TransValues will see the constructor call in order for that call to be transformed into factory method invocation.

This pull request has now been integrated.

Changeset: 8b5c17ba
Author:    Srikanth Adayapalam <sadayapalam at openjdk.org>
URL:       https://git.openjdk.java.net/valhalla/commit/8b5c17ba3d714adaeb43a62624db800122529ed2
Stats:     72 lines in 2 files changed: 71 ins; 0 del; 1 mod

8281166: [lworld] javac should generate BSM to invoke the static factory for value class

-------------

PR: https://git.openjdk.java.net/valhalla/pull/629

From sadayapalam at openjdk.java.net  Mon Feb  7 12:40:46 2022
From: sadayapalam at openjdk.java.net (Srikanth Adayapalam)
Date: Mon, 7 Feb 2022 12:40:46 GMT
Subject: [lworld] RFR: 8281336: [lworld] Remove workaround in LambdaToMethod to
 circumvent BootstrapMethodError
Message-ID: <B8DNUqGZYgKpR77wyjcNVgNR6oG0LFBK6fI9mpPsz1E=.ca796ce0-ca03-4632-9dab-d31fd66c0e8a@github.com>

Remove the workaround in Javac and include the fix from Mandy for runtime in lieu.

-------------

Commit messages:
 - 8281336: Remove workaround in LambdaToMethod to circumvent BootstrapMethodError

Changes: https://git.openjdk.java.net/valhalla/pull/631/files
 Webrev: https://webrevs.openjdk.java.net/?repo=valhalla&pr=631&range=00
  Issue: https://bugs.openjdk.java.net/browse/JDK-8281336
  Stats: 12 lines in 2 files changed: 0 ins; 9 del; 3 mod
  Patch: https://git.openjdk.java.net/valhalla/pull/631.diff
  Fetch: git fetch https://git.openjdk.java.net/valhalla pull/631/head:pull/631

PR: https://git.openjdk.java.net/valhalla/pull/631

From hseigel at openjdk.java.net  Mon Feb  7 13:45:52 2022
From: hseigel at openjdk.java.net (Harold Seigel)
Date: Mon, 7 Feb 2022 13:45:52 GMT
Subject: [lworld] RFR: 8281279: [lworld] Add JVM support for ACC_PERMITS_VALUE
Message-ID: <Bx7Aqv2sFJ4TMvasfAyOIfxdc661Pm23M1GCzy6nFGQ=.3abe36fd-1b91-4ac9-82ed-e67ad16d74f7@github.com>

Please review this fix to add JVM support for ACC_PERMITS_VALUE.  This change throws an ICCE exception when loading a value class whose super class is not java.lang.Object or does not have ACC_PERMITS_VALUE set.  The fix was tested with Mach5 tiers 1-2 on Linux, Mac OS, and Windows and Mach5 tiers 3-5 on Linux x64.

Thanks, Harold

-------------

Commit messages:
 - 8281279: [lworld] Add JVM support for ACC_PERMITS_VALUE

Changes: https://git.openjdk.java.net/valhalla/pull/632/files
 Webrev: https://webrevs.openjdk.java.net/?repo=valhalla&pr=632&range=00
  Issue: https://bugs.openjdk.java.net/browse/JDK-8281279
  Stats: 1659 lines in 15 files changed: 1361 ins; 205 del; 93 mod
  Patch: https://git.openjdk.java.net/valhalla/pull/632.diff
  Fetch: git fetch https://git.openjdk.java.net/valhalla pull/632/head:pull/632

PR: https://git.openjdk.java.net/valhalla/pull/632

From hseigel at openjdk.java.net  Mon Feb  7 15:01:14 2022
From: hseigel at openjdk.java.net (Harold Seigel)
Date: Mon, 7 Feb 2022 15:01:14 GMT
Subject: [lworld] RFR: 8281279: [lworld] Add JVM support for
 ACC_PERMITS_VALUE [v2]
In-Reply-To: <Bx7Aqv2sFJ4TMvasfAyOIfxdc661Pm23M1GCzy6nFGQ=.3abe36fd-1b91-4ac9-82ed-e67ad16d74f7@github.com>
References: <Bx7Aqv2sFJ4TMvasfAyOIfxdc661Pm23M1GCzy6nFGQ=.3abe36fd-1b91-4ac9-82ed-e67ad16d74f7@github.com>
Message-ID: <dRK7Mw3uURcR9H3evNTI_8sfZjvyWv6cU4wN4-8-AUE=.e2e0ae5e-ad99-412a-84c1-f7ee8464446f@github.com>

> Please review this fix to add JVM support for ACC_PERMITS_VALUE.  This change throws an ICCE exception when loading a value class whose super class is not java.lang.Object or does not have ACC_PERMITS_VALUE set.  The fix was tested with Mach5 tiers 1-2 on Linux, Mac OS, and Windows and Mach5 tiers 3-5 on Linux x64.
> 
> Thanks, Harold

Harold Seigel has updated the pull request with a new target base due to a merge or a rebase. The pull request now contains two commits:

 - Merge lworld
 - 8281279: [lworld] Add JVM support for ACC_PERMITS_VALUE

-------------

Changes: https://git.openjdk.java.net/valhalla/pull/632/files
 Webrev: https://webrevs.openjdk.java.net/?repo=valhalla&pr=632&range=01
  Stats: 1659 lines in 15 files changed: 1361 ins; 205 del; 93 mod
  Patch: https://git.openjdk.java.net/valhalla/pull/632.diff
  Fetch: git fetch https://git.openjdk.java.net/valhalla pull/632/head:pull/632

PR: https://git.openjdk.java.net/valhalla/pull/632

From hseigel at openjdk.java.net  Mon Feb  7 15:15:13 2022
From: hseigel at openjdk.java.net (Harold Seigel)
Date: Mon, 7 Feb 2022 15:15:13 GMT
Subject: [lworld] RFR: 8281279: [lworld] Add JVM support for
 ACC_PERMITS_VALUE [v3]
In-Reply-To: <Bx7Aqv2sFJ4TMvasfAyOIfxdc661Pm23M1GCzy6nFGQ=.3abe36fd-1b91-4ac9-82ed-e67ad16d74f7@github.com>
References: <Bx7Aqv2sFJ4TMvasfAyOIfxdc661Pm23M1GCzy6nFGQ=.3abe36fd-1b91-4ac9-82ed-e67ad16d74f7@github.com>
Message-ID: <4q-fdSLg8oc_fR5c5Qt0Fh7cfm37omb6Qi944drCzaE=.474f196d-78d1-4d36-80e6-e3c61bdbab04@github.com>

> Please review this fix to add JVM support for ACC_PERMITS_VALUE.  This change throws an ICCE exception when loading a value class whose super class is not java.lang.Object or does not have ACC_PERMITS_VALUE set.  The fix was tested with Mach5 tiers 1-2 on Linux, Mac OS, and Windows and Mach5 tiers 3-5 on Linux x64.
> 
> Thanks, Harold

Harold Seigel has updated the pull request incrementally with one additional commit since the last revision:

  Update test TestResolvedJavaType.java

-------------

Changes:
  - all: https://git.openjdk.java.net/valhalla/pull/632/files
  - new: https://git.openjdk.java.net/valhalla/pull/632/files/c2cbdfe0..81dc07be

Webrevs:
 - full: https://webrevs.openjdk.java.net/?repo=valhalla&pr=632&range=02
 - incr: https://webrevs.openjdk.java.net/?repo=valhalla&pr=632&range=01-02

  Stats: 4 lines in 1 file changed: 0 ins; 0 del; 4 mod
  Patch: https://git.openjdk.java.net/valhalla/pull/632.diff
  Fetch: git fetch https://git.openjdk.java.net/valhalla pull/632/head:pull/632

PR: https://git.openjdk.java.net/valhalla/pull/632

From thartmann at openjdk.java.net  Mon Feb  7 15:23:38 2022
From: thartmann at openjdk.java.net (Tobias Hartmann)
Date: Mon, 7 Feb 2022 15:23:38 GMT
Subject: [lworld] RFR: 8281325: [lworld] Unused code emitted for unpacking
 arguments leads to code buffer overflow
Message-ID: <NHJvRUy_dZY02yDvkKbXFZPKIAAF8zYhj0o0YVSutn0=.e7dd8c37-b8bb-44fd-b7fe-6b85ab0652b7@github.com>

An unrelated change to javac that came in with the mainline merge triggered this:
When emitting code to scalarize a buffered inline type argument in the method entries, there is sometimes an overlap with registers/stack slots occupied by another argument that needs to be resolved before we can make progress. We should not emit any code in such a case and proceed with the next argument (we will revisit this argument later, once the conflict has been resolved).

Some of that code will be refactored by [JDK-8278390](https://bugs.openjdk.java.net/browse/JDK-8278390) anyway.

Thanks,
Tobias

-------------

Commit messages:
 - Typo
 - Removed from problem list
 - 8281325: [lworld] Unused code emitted for unpacking arguments leads to code buffer overflow

Changes: https://git.openjdk.java.net/valhalla/pull/633/files
 Webrev: https://webrevs.openjdk.java.net/?repo=valhalla&pr=633&range=00
  Issue: https://bugs.openjdk.java.net/browse/JDK-8281325
  Stats: 64 lines in 3 files changed: 31 ins; 20 del; 13 mod
  Patch: https://git.openjdk.java.net/valhalla/pull/633.diff
  Fetch: git fetch https://git.openjdk.java.net/valhalla pull/633/head:pull/633

PR: https://git.openjdk.java.net/valhalla/pull/633

From sadayapalam at openjdk.java.net  Mon Feb  7 15:26:25 2022
From: sadayapalam at openjdk.java.net (Srikanth Adayapalam)
Date: Mon, 7 Feb 2022 15:26:25 GMT
Subject: [lworld] RFR: 8280194: Abstract classes that allow value
 subclasses should be marked ACC_PERMITS_VALUE (0x0040) [v2]
In-Reply-To: <Q4ghm3RJlgzwFYN9we3znt5tSdwnA2o79qbEy-v0Fv4=.49d0e22d-70c7-4287-b29c-46859088ec65@github.com>
References: <6R7uyjsabNXiT2YljFGI3ohOzBDtOTIYMF-MdK6tncI=.4ff4eb00-4cdc-4123-887c-a827cc4dc35b@github.com>
 <Q4ghm3RJlgzwFYN9we3znt5tSdwnA2o79qbEy-v0Fv4=.49d0e22d-70c7-4287-b29c-46859088ec65@github.com>
Message-ID: <56yPDp_p_oWbgGV656OTpnEZ8Det3AXr7aa-alK4MD0=.55810b18-0ea0-4387-8fe1-c33ece5f96c3@github.com>

On Tue, 1 Feb 2022 11:33:23 GMT, Srikanth Adayapalam <sadayapalam at openjdk.org> wrote:

>> Mark abstract classes that qualify with ACC_PERMITS_VALUE flag.
>
> Srikanth Adayapalam has updated the pull request incrementally with one additional commit since the last revision:
> 
>   Incorporate review comments

I took a look at the Modifier.getString enhancement - at the moment ACC_PERMITS_VALUE is not surfaced at all in these tests. So we can leave it as is for the time being.

-------------

PR: https://git.openjdk.java.net/valhalla/pull/622

From thartmann at openjdk.java.net  Mon Feb  7 15:50:42 2022
From: thartmann at openjdk.java.net (Tobias Hartmann)
Date: Mon, 7 Feb 2022 15:50:42 GMT
Subject: [lworld] Integrated: 8281325: [lworld] Unused code emitted for
 unpacking arguments leads to code buffer overflow
In-Reply-To: <NHJvRUy_dZY02yDvkKbXFZPKIAAF8zYhj0o0YVSutn0=.e7dd8c37-b8bb-44fd-b7fe-6b85ab0652b7@github.com>
References: <NHJvRUy_dZY02yDvkKbXFZPKIAAF8zYhj0o0YVSutn0=.e7dd8c37-b8bb-44fd-b7fe-6b85ab0652b7@github.com>
Message-ID: <PfsD2_9f0PF24sqhNc_-BFZEGzha4yvTVX2-dt2nv4A=.3be9c4e7-b580-477d-afab-b4d547ba1e40@github.com>

On Mon, 7 Feb 2022 15:17:05 GMT, Tobias Hartmann <thartmann at openjdk.org> wrote:

> An unrelated change to javac that came in with the mainline merge triggered this:
> When emitting code to scalarize a buffered inline type argument in the method entries, there is sometimes an overlap with registers/stack slots occupied by another argument that needs to be resolved before we can make progress. We should not emit any code in such a case and proceed with the next argument (we will revisit this argument later, once the conflict has been resolved).
> 
> Some of that code will be refactored by [JDK-8278390](https://bugs.openjdk.java.net/browse/JDK-8278390) anyway.
> 
> Thanks,
> Tobias

This pull request has now been integrated.

Changeset: 2f2fa459
Author:    Tobias Hartmann <thartmann at openjdk.org>
URL:       https://git.openjdk.java.net/valhalla/commit/2f2fa459af4215085b8bf78be7fc11619a250b59
Stats:     64 lines in 3 files changed: 31 ins; 20 del; 13 mod

8281325: [lworld] Unused code emitted for unpacking arguments leads to code buffer overflow

-------------

PR: https://git.openjdk.java.net/valhalla/pull/633

From hseigel at openjdk.java.net  Mon Feb  7 18:05:13 2022
From: hseigel at openjdk.java.net (Harold Seigel)
Date: Mon, 7 Feb 2022 18:05:13 GMT
Subject: [lworld] RFR: 8281279: [lworld] Add JVM support for
 ACC_PERMITS_VALUE [v4]
In-Reply-To: <Bx7Aqv2sFJ4TMvasfAyOIfxdc661Pm23M1GCzy6nFGQ=.3abe36fd-1b91-4ac9-82ed-e67ad16d74f7@github.com>
References: <Bx7Aqv2sFJ4TMvasfAyOIfxdc661Pm23M1GCzy6nFGQ=.3abe36fd-1b91-4ac9-82ed-e67ad16d74f7@github.com>
Message-ID: <M1l0Ipkd9yue9JsyYSkBN9SC-DEdotCpIY8pG_2XLHU=.ad94ab17-4b2a-42aa-80ec-5645b3b1ea72@github.com>

> Please review this fix to add JVM support for ACC_PERMITS_VALUE.  This change throws an ICCE exception when loading a value class whose super class is not java.lang.Object or does not have ACC_PERMITS_VALUE set.  The fix was tested with Mach5 tiers 1-2 on Linux, Mac OS, and Windows and Mach5 tiers 3-5 on Linux x64.
> 
> Thanks, Harold

Harold Seigel has updated the pull request incrementally with one additional commit since the last revision:

  some additional fixes

-------------

Changes:
  - all: https://git.openjdk.java.net/valhalla/pull/632/files
  - new: https://git.openjdk.java.net/valhalla/pull/632/files/81dc07be..71b99f48

Webrevs:
 - full: https://webrevs.openjdk.java.net/?repo=valhalla&pr=632&range=03
 - incr: https://webrevs.openjdk.java.net/?repo=valhalla&pr=632&range=02-03

  Stats: 9 lines in 3 files changed: 2 ins; 0 del; 7 mod
  Patch: https://git.openjdk.java.net/valhalla/pull/632.diff
  Fetch: git fetch https://git.openjdk.java.net/valhalla pull/632/head:pull/632

PR: https://git.openjdk.java.net/valhalla/pull/632

From fparain at openjdk.java.net  Mon Feb  7 18:05:14 2022
From: fparain at openjdk.java.net (Frederic Parain)
Date: Mon, 7 Feb 2022 18:05:14 GMT
Subject: [lworld] RFR: 8281279: [lworld] Add JVM support for
 ACC_PERMITS_VALUE [v3]
In-Reply-To: <4q-fdSLg8oc_fR5c5Qt0Fh7cfm37omb6Qi944drCzaE=.474f196d-78d1-4d36-80e6-e3c61bdbab04@github.com>
References: <Bx7Aqv2sFJ4TMvasfAyOIfxdc661Pm23M1GCzy6nFGQ=.3abe36fd-1b91-4ac9-82ed-e67ad16d74f7@github.com>
 <4q-fdSLg8oc_fR5c5Qt0Fh7cfm37omb6Qi944drCzaE=.474f196d-78d1-4d36-80e6-e3c61bdbab04@github.com>
Message-ID: <B9Mg-0cS2U8WFcymTO-tvhItjAhopB3vNRGgOPSr_-8=.fda24546-ae0e-475b-9ae9-884b0134ac41@github.com>

On Mon, 7 Feb 2022 15:15:13 GMT, Harold Seigel <hseigel at openjdk.org> wrote:

>> Please review this fix to add JVM support for ACC_PERMITS_VALUE.  This change throws an ICCE exception when loading a value class whose super class is not java.lang.Object or does not have ACC_PERMITS_VALUE set.  The fix was tested with Mach5 tiers 1-2 on Linux, Mac OS, and Windows and Mach5 tiers 3-5 on Linux x64.
>> 
>> Thanks, Harold
>
> Harold Seigel has updated the pull request incrementally with one additional commit since the last revision:
> 
>   Update test TestResolvedJavaType.java

src/hotspot/share/classfile/classFileParser.cpp line 4596:

> 4594:       return;
> 4595:     }
> 4596: 

This is not the behavior written in the JVMS draft for value classes, which requires the direct super class of a value class to have the ACC_PERMITS _VALUE flag set. However, because we haven't changed java.lang.Object into an abstract class, java.lang.Object cannot have the ACC_PERMITS _VALUE flag set. Currently, I see no other solution than the one you have implemented, adding a special case when the direct super class is java.lang.Object. But it would be useful to add a comment in the source code and the CR to remember and track this discrepancy.

-------------

PR: https://git.openjdk.java.net/valhalla/pull/632

From mchung at openjdk.java.net  Mon Feb  7 18:11:39 2022
From: mchung at openjdk.java.net (Mandy Chung)
Date: Mon, 7 Feb 2022 18:11:39 GMT
Subject: [lworld] RFR: 8281336: [lworld] Remove workaround in
 LambdaToMethod to circumvent BootstrapMethodError
In-Reply-To: <B8DNUqGZYgKpR77wyjcNVgNR6oG0LFBK6fI9mpPsz1E=.ca796ce0-ca03-4632-9dab-d31fd66c0e8a@github.com>
References: <B8DNUqGZYgKpR77wyjcNVgNR6oG0LFBK6fI9mpPsz1E=.ca796ce0-ca03-4632-9dab-d31fd66c0e8a@github.com>
Message-ID: <kzzsCW1NMA-0BulHKGUalb1YotvIyPHCy4XBJfOwEhg=.b2c7ff53-a0cc-4c6e-a2ab-3edcacf4295a@github.com>

On Mon, 7 Feb 2022 12:33:39 GMT, Srikanth Adayapalam <sadayapalam at openjdk.org> wrote:

> Remove the workaround in Javac and include the fix from Mandy for runtime in lieu.

The runtime change looks fine.  Is there an existing test verifying this?

-------------

PR: https://git.openjdk.java.net/valhalla/pull/631

From hseigel at openjdk.java.net  Mon Feb  7 18:15:54 2022
From: hseigel at openjdk.java.net (Harold Seigel)
Date: Mon, 7 Feb 2022 18:15:54 GMT
Subject: [lworld] RFR: 8281279: [lworld] Add JVM support for
 ACC_PERMITS_VALUE [v5]
In-Reply-To: <Bx7Aqv2sFJ4TMvasfAyOIfxdc661Pm23M1GCzy6nFGQ=.3abe36fd-1b91-4ac9-82ed-e67ad16d74f7@github.com>
References: <Bx7Aqv2sFJ4TMvasfAyOIfxdc661Pm23M1GCzy6nFGQ=.3abe36fd-1b91-4ac9-82ed-e67ad16d74f7@github.com>
Message-ID: <txivvMsAReynqMQjpKRvJffLaiEjfi1ZgCfj9zZy-7Y=.1afa863d-f2c3-47f2-a9e2-d916e3beaefb@github.com>

> Please review this fix to add JVM support for ACC_PERMITS_VALUE.  This change throws an ICCE exception when loading a value class whose super class is not java.lang.Object or does not have ACC_PERMITS_VALUE set.  The fix was tested with Mach5 tiers 1-2 on Linux, Mac OS, and Windows and Mach5 tiers 3-5 on Linux x64.
> 
> Thanks, Harold

Harold Seigel has updated the pull request incrementally with one additional commit since the last revision:

  add j.l.Object comment

-------------

Changes:
  - all: https://git.openjdk.java.net/valhalla/pull/632/files
  - new: https://git.openjdk.java.net/valhalla/pull/632/files/71b99f48..32609d17

Webrevs:
 - full: https://webrevs.openjdk.java.net/?repo=valhalla&pr=632&range=04
 - incr: https://webrevs.openjdk.java.net/?repo=valhalla&pr=632&range=03-04

  Stats: 5 lines in 1 file changed: 4 ins; 0 del; 1 mod
  Patch: https://git.openjdk.java.net/valhalla/pull/632.diff
  Fetch: git fetch https://git.openjdk.java.net/valhalla pull/632/head:pull/632

PR: https://git.openjdk.java.net/valhalla/pull/632

From hseigel at openjdk.java.net  Mon Feb  7 18:15:56 2022
From: hseigel at openjdk.java.net (Harold Seigel)
Date: Mon, 7 Feb 2022 18:15:56 GMT
Subject: [lworld] RFR: 8281279: [lworld] Add JVM support for
 ACC_PERMITS_VALUE [v3]
In-Reply-To: <B9Mg-0cS2U8WFcymTO-tvhItjAhopB3vNRGgOPSr_-8=.fda24546-ae0e-475b-9ae9-884b0134ac41@github.com>
References: <Bx7Aqv2sFJ4TMvasfAyOIfxdc661Pm23M1GCzy6nFGQ=.3abe36fd-1b91-4ac9-82ed-e67ad16d74f7@github.com>
 <4q-fdSLg8oc_fR5c5Qt0Fh7cfm37omb6Qi944drCzaE=.474f196d-78d1-4d36-80e6-e3c61bdbab04@github.com>
 <B9Mg-0cS2U8WFcymTO-tvhItjAhopB3vNRGgOPSr_-8=.fda24546-ae0e-475b-9ae9-884b0134ac41@github.com>
Message-ID: <7Ar4JIVs4FX3JhqUfykv5uZlavzia_VXwOYvCtjtppM=.1e83938e-ff8a-4f4f-b800-2c286584b61b@github.com>

On Mon, 7 Feb 2022 17:59:25 GMT, Frederic Parain <fparain at openjdk.org> wrote:

>> Harold Seigel has updated the pull request incrementally with one additional commit since the last revision:
>> 
>>   Update test TestResolvedJavaType.java
>
> src/hotspot/share/classfile/classFileParser.cpp line 4596:
> 
>> 4594:       return;
>> 4595:     }
>> 4596: 
> 
> This is not the behavior written in the JVMS draft for value classes, which requires the direct super class of a value class to have the ACC_PERMITS _VALUE flag set. However, because we haven't changed java.lang.Object into an abstract class, java.lang.Object cannot have the ACC_PERMITS _VALUE flag set. Currently, I see no other solution than the one you have implemented, adding a special case when the direct super class is java.lang.Object. But it would be useful to add a comment in the source code and the CR to remember and track this discrepancy.

Thanks for suggesting this.  I added a comment.

-------------

PR: https://git.openjdk.java.net/valhalla/pull/632

From duke at openjdk.java.net  Mon Feb  7 22:45:25 2022
From: duke at openjdk.java.net (Aggelos Biboudis)
Date: Mon, 7 Feb 2022 22:45:25 GMT
Subject: [lworld] RFR: 8281336: [lworld] Remove workaround in
 LambdaToMethod to circumvent BootstrapMethodError
In-Reply-To: <B8DNUqGZYgKpR77wyjcNVgNR6oG0LFBK6fI9mpPsz1E=.ca796ce0-ca03-4632-9dab-d31fd66c0e8a@github.com>
References: <B8DNUqGZYgKpR77wyjcNVgNR6oG0LFBK6fI9mpPsz1E=.ca796ce0-ca03-4632-9dab-d31fd66c0e8a@github.com>
Message-ID: <l18xGzmymQBLR2pD83K3hKMQ0XFegaUy4_hBKuqCDfQ=.c0120c15-0e2a-4975-8744-b10457c4e429@github.com>

On Mon, 7 Feb 2022 12:33:39 GMT, Srikanth Adayapalam <sadayapalam at openjdk.org> wrote:

> Remove the workaround in Javac and include the fix from Mandy for runtime in lieu.

Maybe we can use the test from the the original PR?


import java.util.function.Supplier;

public primitive class CheckRefLambda {
    int theInteger;

    CheckRefLambda(int newValue){
        theInteger=newValue;
    }

    public Supplier<CheckRefLambda.ref> makeCopier(){
        return()->new CheckRefLambda(this.theInteger + 1);
    }

    public static void main(String[]args){
        CheckRefLambda rec=new CheckRefLambda(42);
        CheckRefLambda rec2=rec.makeCopier().get();
        if(rec2!=new CheckRefLambda(43)){
            throw new AssertionError("Classes should be equal");
        }
    }
}

-------------

PR: https://git.openjdk.java.net/valhalla/pull/631

From sadayapalam at openjdk.java.net  Tue Feb  8 00:14:25 2022
From: sadayapalam at openjdk.java.net (Srikanth Adayapalam)
Date: Tue, 8 Feb 2022 00:14:25 GMT
Subject: [lworld] RFR: 8281336: [lworld] Remove workaround in
 LambdaToMethod to circumvent BootstrapMethodError
In-Reply-To: <kzzsCW1NMA-0BulHKGUalb1YotvIyPHCy4XBJfOwEhg=.b2c7ff53-a0cc-4c6e-a2ab-3edcacf4295a@github.com>
References: <B8DNUqGZYgKpR77wyjcNVgNR6oG0LFBK6fI9mpPsz1E=.ca796ce0-ca03-4632-9dab-d31fd66c0e8a@github.com>
 <kzzsCW1NMA-0BulHKGUalb1YotvIyPHCy4XBJfOwEhg=.b2c7ff53-a0cc-4c6e-a2ab-3edcacf4295a@github.com>
Message-ID: <hZH_pFp578TVGPxWdu4vbRlLtCXCeEk3aOj4qJIOTZA=.4553d845-3cd7-4277-a064-61689fbd0f02@github.com>

On Mon, 7 Feb 2022 18:08:29 GMT, Mandy Chung <mchung at openjdk.org> wrote:

> The runtime change looks fine. Is there an existing test verifying this?

Yes, indeed. test/langtools/tools/javac/valhalla/lworld-values/StreamsTest.java fails with only javac changes without the runtime changes.

-------------

PR: https://git.openjdk.java.net/valhalla/pull/631

From mchung at openjdk.java.net  Tue Feb  8 00:19:19 2022
From: mchung at openjdk.java.net (Mandy Chung)
Date: Tue, 8 Feb 2022 00:19:19 GMT
Subject: [lworld] RFR: 8281336: [lworld] Remove workaround in
 LambdaToMethod to circumvent BootstrapMethodError
In-Reply-To: <B8DNUqGZYgKpR77wyjcNVgNR6oG0LFBK6fI9mpPsz1E=.ca796ce0-ca03-4632-9dab-d31fd66c0e8a@github.com>
References: <B8DNUqGZYgKpR77wyjcNVgNR6oG0LFBK6fI9mpPsz1E=.ca796ce0-ca03-4632-9dab-d31fd66c0e8a@github.com>
Message-ID: <SckvMbJ8PaWR539SqEx7b0_GhTSNL6Y11j_esq7ihzU=.52480909-808f-4051-b39f-e4bbbabb8d3d@github.com>

On Mon, 7 Feb 2022 12:33:39 GMT, Srikanth Adayapalam <sadayapalam at openjdk.org> wrote:

> Remove the workaround in Javac and include the fix from Mandy for runtime in lieu.

Marked as reviewed by mchung (Committer).

-------------

PR: https://git.openjdk.java.net/valhalla/pull/631

From sadayapalam at openjdk.java.net  Tue Feb  8 00:35:38 2022
From: sadayapalam at openjdk.java.net (Srikanth Adayapalam)
Date: Tue, 8 Feb 2022 00:35:38 GMT
Subject: [lworld] RFR: 8281336: [lworld] Remove workaround in
 LambdaToMethod to circumvent BootstrapMethodError
In-Reply-To: <l18xGzmymQBLR2pD83K3hKMQ0XFegaUy4_hBKuqCDfQ=.c0120c15-0e2a-4975-8744-b10457c4e429@github.com>
References: <B8DNUqGZYgKpR77wyjcNVgNR6oG0LFBK6fI9mpPsz1E=.ca796ce0-ca03-4632-9dab-d31fd66c0e8a@github.com>
 <l18xGzmymQBLR2pD83K3hKMQ0XFegaUy4_hBKuqCDfQ=.c0120c15-0e2a-4975-8744-b10457c4e429@github.com>
Message-ID: <8uadIR33TCdqAIDE-muIL3bjTuZYvQCgIsodQ4_v0JE=.1bdfe57a-b38b-4be7-aef1-f5f4d4a51a73@github.com>

On Mon, 7 Feb 2022 22:42:37 GMT, Aggelos Biboudis <duke at openjdk.java.net> wrote:

> Maybe we can use the test from the the original PR?
> 
> ```
> import java.util.function.Supplier;
> 
> public primitive class CheckRefLambda {
>     int theInteger;
> 
>     CheckRefLambda(int newValue){
>         theInteger=newValue;
>     }
> 
>     public Supplier<CheckRefLambda.ref> makeCopier(){
>         return()->new CheckRefLambda(this.theInteger + 1);
>     }
> 
>     public static void main(String[]args){
>         CheckRefLambda rec=new CheckRefLambda(42);
>         CheckRefLambda rec2=rec.makeCopier().get();
>         if(rec2!=new CheckRefLambda(43)){
>             throw new AssertionError("Classes should be equal");
>         }
>     }
> }
> ```

Much water has flown under the bridge - we have ripped out all support for ref-default classes and so the meanings of unadorned/naked names have changed. (in the original scenario the name CheckRefLambda is a primitive reference but in what you have put together, CheckRefLambda is an instance of a primitive class) This test case does not fail anymore without any changes (javac and/or runtime). Since we already have a javac test covering the PR, we should be good I would say.

-------------

PR: https://git.openjdk.java.net/valhalla/pull/631

From sadayapalam at openjdk.java.net  Tue Feb  8 00:35:39 2022
From: sadayapalam at openjdk.java.net (Srikanth Adayapalam)
Date: Tue, 8 Feb 2022 00:35:39 GMT
Subject: [lworld] Integrated: 8281336: [lworld] Remove workaround in
 LambdaToMethod to circumvent BootstrapMethodError
In-Reply-To: <B8DNUqGZYgKpR77wyjcNVgNR6oG0LFBK6fI9mpPsz1E=.ca796ce0-ca03-4632-9dab-d31fd66c0e8a@github.com>
References: <B8DNUqGZYgKpR77wyjcNVgNR6oG0LFBK6fI9mpPsz1E=.ca796ce0-ca03-4632-9dab-d31fd66c0e8a@github.com>
Message-ID: <H50UwxCGbVxWEzDdY_Zl-rfzfnRuIA5LY4_UyGT5348=.ffcb83ed-2506-44cf-af82-09d9276303ed@github.com>

On Mon, 7 Feb 2022 12:33:39 GMT, Srikanth Adayapalam <sadayapalam at openjdk.org> wrote:

> Remove the workaround in Javac and include the fix from Mandy for runtime in lieu.

This pull request has now been integrated.

Changeset: 2571b270
Author:    Srikanth Adayapalam <sadayapalam at openjdk.org>
URL:       https://git.openjdk.java.net/valhalla/commit/2571b270de06c209d5bfdb86a1a1929e6aa711b4
Stats:     12 lines in 2 files changed: 0 ins; 9 del; 3 mod

8281336: [lworld] Remove workaround in LambdaToMethod to circumvent BootstrapMethodError
8274399: [lworld] LambdaConversionException thrown when the receiver type is a primitive reference type and the implementation type is the primitive value type of the same class

Co-authored-by: Mandy Chung <mchung at openjdk.org>
Reviewed-by: mchung

-------------

PR: https://git.openjdk.java.net/valhalla/pull/631

From sadayapalam at openjdk.java.net  Tue Feb  8 06:41:06 2022
From: sadayapalam at openjdk.java.net (Srikanth Adayapalam)
Date: Tue, 8 Feb 2022 06:41:06 GMT
Subject: [lworld] RFR: 8281013: Errors.PrimitiveClassDoesNotSupport diagnostic
 should apply to value classes also
Message-ID: <O2alphURtAT3CtuGNQEVaYphQj2pF4U3XcQv6UGFePQ=.243e2457-1df2-4bad-b75c-cad6d8500ffa@github.com>

Diagnose value classes accessing various jlO methods that don't work with B2/B3 classes

-------------

Commit messages:
 - 8281013: Errors.PrimitiveClassDoesNotSupport diagnostic should apply to value classes also

Changes: https://git.openjdk.java.net/valhalla/pull/635/files
 Webrev: https://webrevs.openjdk.java.net/?repo=valhalla&pr=635&range=00
  Issue: https://bugs.openjdk.java.net/browse/JDK-8281013
  Stats: 87 lines in 9 files changed: 50 ins; 0 del; 37 mod
  Patch: https://git.openjdk.java.net/valhalla/pull/635.diff
  Fetch: git fetch https://git.openjdk.java.net/valhalla pull/635/head:pull/635

PR: https://git.openjdk.java.net/valhalla/pull/635

From duke at openjdk.java.net  Tue Feb  8 10:25:42 2022
From: duke at openjdk.java.net (Aggelos Biboudis)
Date: Tue, 8 Feb 2022 10:25:42 GMT
Subject: [lworld] RFR: 8281013: Errors.PrimitiveClassDoesNotSupport
 diagnostic should apply to value classes also
In-Reply-To: <O2alphURtAT3CtuGNQEVaYphQj2pF4U3XcQv6UGFePQ=.243e2457-1df2-4bad-b75c-cad6d8500ffa@github.com>
References: <O2alphURtAT3CtuGNQEVaYphQj2pF4U3XcQv6UGFePQ=.243e2457-1df2-4bad-b75c-cad6d8500ffa@github.com>
Message-ID: <XL2WP7B_o7z_BWryIL9l0Vj2of1TAkK-lmvRLCFGYVI=.a7d9a57b-886b-43eb-a6b6-ce2627ecf52c@github.com>

On Tue, 8 Feb 2022 06:32:17 GMT, Srikanth Adayapalam <sadayapalam at openjdk.org> wrote:

> Diagnose value classes accessing various jlO methods that don't work with B2/B3 classes

Looks good! Should we reference 8237066 as well in the PR for administrative purposes? (I noticed it is referred in the JTREG comment).

-------------

Marked as reviewed by biboudis at github.com (no known OpenJDK username).

PR: https://git.openjdk.java.net/valhalla/pull/635

From sadayapalam at openjdk.java.net  Tue Feb  8 10:34:41 2022
From: sadayapalam at openjdk.java.net (Srikanth Adayapalam)
Date: Tue, 8 Feb 2022 10:34:41 GMT
Subject: [lworld] RFR: 8281013: Errors.PrimitiveClassDoesNotSupport
 diagnostic should apply to value classes also
In-Reply-To: <XL2WP7B_o7z_BWryIL9l0Vj2of1TAkK-lmvRLCFGYVI=.a7d9a57b-886b-43eb-a6b6-ce2627ecf52c@github.com>
References: <O2alphURtAT3CtuGNQEVaYphQj2pF4U3XcQv6UGFePQ=.243e2457-1df2-4bad-b75c-cad6d8500ffa@github.com>
 <XL2WP7B_o7z_BWryIL9l0Vj2of1TAkK-lmvRLCFGYVI=.a7d9a57b-886b-43eb-a6b6-ce2627ecf52c@github.com>
Message-ID: <K_e2lZfeGsWdlKfSCvJYs7LaNhbCw5ITcPYqdpvjlWw=.bb37675e-36bf-4df9-8674-531da82fb6cc@github.com>

On Tue, 8 Feb 2022 10:22:07 GMT, Aggelos Biboudis <duke at openjdk.java.net> wrote:

> Looks good! Should we reference 8237066 as well in the PR for administrative purposes? (I noticed it is referred in the JTREG comment).

The new test is a starts off with existing test which was introduced for JDK-8237066 - As a part of the testsuite cleanup much of these references should really be cleaned up. I am not sure these bug ids would  make sense during and post integration to mainline JDK - these are pre-release bugs. 

Referencing 8237066 in the PR - I don't think something like /issue add 8237066 would work - since that would look for a open ticket that should also be closed alongside.

-------------

PR: https://git.openjdk.java.net/valhalla/pull/635

From sadayapalam at openjdk.java.net  Tue Feb  8 10:44:35 2022
From: sadayapalam at openjdk.java.net (Srikanth Adayapalam)
Date: Tue, 8 Feb 2022 10:44:35 GMT
Subject: [lworld] Integrated: 8281013: Errors.PrimitiveClassDoesNotSupport
 diagnostic should apply to value classes also
In-Reply-To: <O2alphURtAT3CtuGNQEVaYphQj2pF4U3XcQv6UGFePQ=.243e2457-1df2-4bad-b75c-cad6d8500ffa@github.com>
References: <O2alphURtAT3CtuGNQEVaYphQj2pF4U3XcQv6UGFePQ=.243e2457-1df2-4bad-b75c-cad6d8500ffa@github.com>
Message-ID: <ToQ_0rfcmrhK8RB_664VOhszVAObwIXl8_pBdVsLF18=.dc01a5b1-af59-4f78-9926-c7d3be58fdb6@github.com>

On Tue, 8 Feb 2022 06:32:17 GMT, Srikanth Adayapalam <sadayapalam at openjdk.org> wrote:

> Diagnose value classes accessing various jlO methods that don't work with B2/B3 classes

This pull request has now been integrated.

Changeset: fc22b797
Author:    Srikanth Adayapalam <sadayapalam at openjdk.org>
URL:       https://git.openjdk.java.net/valhalla/commit/fc22b797a7438dec5dbae8bb0f536207b5619765
Stats:     87 lines in 9 files changed: 50 ins; 0 del; 37 mod

8281013: Errors.PrimitiveClassDoesNotSupport diagnostic should apply to value classes also

-------------

PR: https://git.openjdk.java.net/valhalla/pull/635

From sadayapalam at openjdk.java.net  Tue Feb  8 12:49:47 2022
From: sadayapalam at openjdk.java.net (Srikanth Adayapalam)
Date: Tue, 8 Feb 2022 12:49:47 GMT
Subject: [lworld] RFR: 8281026: Allow for
 compiler.note.cant.instantiate.object.directly to be suppressed via an option
Message-ID: <76skw4CeVr0i6WFPBK4ToC3eQc661eq_0QKi8TjKob0=.ea4ad890-0e7f-4357-a4fb-dc85cb0de50e@github.com>

Introduce the javac option -XDtolerateObjectInstantiation to suppress the warning.

-------------

Commit messages:
 - 8281026: Allow for compiler.note.cant.instantiate.object.directly to be suppressed via an option

Changes: https://git.openjdk.java.net/valhalla/pull/636/files
 Webrev: https://webrevs.openjdk.java.net/?repo=valhalla&pr=636&range=00
  Issue: https://bugs.openjdk.java.net/browse/JDK-8281026
  Stats: 20 lines in 4 files changed: 14 ins; 0 del; 6 mod
  Patch: https://git.openjdk.java.net/valhalla/pull/636.diff
  Fetch: git fetch https://git.openjdk.java.net/valhalla pull/636/head:pull/636

PR: https://git.openjdk.java.net/valhalla/pull/636

From roland at openjdk.java.net  Tue Feb  8 13:45:45 2022
From: roland at openjdk.java.net (Roland Westrelin)
Date: Tue, 8 Feb 2022 13:45:45 GMT
Subject: [lworld] RFR: 8281378: [lworld] Crash in
 LateInlineMHCallGenerator::do_late_inline_check
Message-ID: <E4kOCNEYmR9VUwx4k7HaujUWNkSPMfOJfvHAvHcBE14=.96187beb-d63f-4166-8b05-16e26368f7a2@github.com>

I think this is caused by the support that was added recently for
virtual call late inlining and I propose tweaking the
AlwaysIncrementalInline logic for that case.

-------------

Commit messages:
 - fix

Changes: https://git.openjdk.java.net/valhalla/pull/637/files
 Webrev: https://webrevs.openjdk.java.net/?repo=valhalla&pr=637&range=00
  Issue: https://bugs.openjdk.java.net/browse/JDK-8281378
  Stats: 3 lines in 1 file changed: 1 ins; 0 del; 2 mod
  Patch: https://git.openjdk.java.net/valhalla/pull/637.diff
  Fetch: git fetch https://git.openjdk.java.net/valhalla pull/637/head:pull/637

PR: https://git.openjdk.java.net/valhalla/pull/637

From thartmann at openjdk.java.net  Tue Feb  8 14:02:27 2022
From: thartmann at openjdk.java.net (Tobias Hartmann)
Date: Tue, 8 Feb 2022 14:02:27 GMT
Subject: [lworld] RFR: 8281378: [lworld] Crash in
 LateInlineMHCallGenerator::do_late_inline_check
In-Reply-To: <E4kOCNEYmR9VUwx4k7HaujUWNkSPMfOJfvHAvHcBE14=.96187beb-d63f-4166-8b05-16e26368f7a2@github.com>
References: <E4kOCNEYmR9VUwx4k7HaujUWNkSPMfOJfvHAvHcBE14=.96187beb-d63f-4166-8b05-16e26368f7a2@github.com>
Message-ID: <PpZc5QXVjU5VdX8XF2cQf_-q8EcmOQzwZLt9iZc5JJA=.2d6a5fdc-2b31-4a0b-91e2-64090433ba32@github.com>

On Tue, 8 Feb 2022 13:40:14 GMT, Roland Westrelin <roland at openjdk.org> wrote:

> I think this is caused by the support that was added recently for
> virtual call late inlining and I propose tweaking the
> AlwaysIncrementalInline logic for that case.

That looks reasonable to me. Thanks for quickly fixing!

-------------

Marked as reviewed by thartmann (Committer).

PR: https://git.openjdk.java.net/valhalla/pull/637

From roland at openjdk.java.net  Tue Feb  8 14:02:28 2022
From: roland at openjdk.java.net (Roland Westrelin)
Date: Tue, 8 Feb 2022 14:02:28 GMT
Subject: [lworld] RFR: 8281378: [lworld] Crash in
 LateInlineMHCallGenerator::do_late_inline_check
In-Reply-To: <PpZc5QXVjU5VdX8XF2cQf_-q8EcmOQzwZLt9iZc5JJA=.2d6a5fdc-2b31-4a0b-91e2-64090433ba32@github.com>
References: <E4kOCNEYmR9VUwx4k7HaujUWNkSPMfOJfvHAvHcBE14=.96187beb-d63f-4166-8b05-16e26368f7a2@github.com>
 <PpZc5QXVjU5VdX8XF2cQf_-q8EcmOQzwZLt9iZc5JJA=.2d6a5fdc-2b31-4a0b-91e2-64090433ba32@github.com>
Message-ID: <UpT1WGFdgetOv8vG7n8uHzfREC_jFHfL-1LbwoXKrpc=.e735df9a-1a4a-44fd-a3e5-725ba3a44827@github.com>

On Tue, 8 Feb 2022 13:56:20 GMT, Tobias Hartmann <thartmann at openjdk.org> wrote:

>> I think this is caused by the support that was added recently for
>> virtual call late inlining and I propose tweaking the
>> AlwaysIncrementalInline logic for that case.
>
> That looks reasonable to me. Thanks for quickly fixing!

@TobiHartmann thanks for the review

-------------

PR: https://git.openjdk.java.net/valhalla/pull/637

From roland at openjdk.java.net  Tue Feb  8 14:02:29 2022
From: roland at openjdk.java.net (Roland Westrelin)
Date: Tue, 8 Feb 2022 14:02:29 GMT
Subject: [lworld] Integrated: 8281378: [lworld] Crash in
 LateInlineMHCallGenerator::do_late_inline_check
In-Reply-To: <E4kOCNEYmR9VUwx4k7HaujUWNkSPMfOJfvHAvHcBE14=.96187beb-d63f-4166-8b05-16e26368f7a2@github.com>
References: <E4kOCNEYmR9VUwx4k7HaujUWNkSPMfOJfvHAvHcBE14=.96187beb-d63f-4166-8b05-16e26368f7a2@github.com>
Message-ID: <NSN7p6Fs1TUz9UEi03wEOr9q1-xGEgkF9e6eCNeHNTA=.76e4a253-dd47-48b9-84fb-fc8781a2fe49@github.com>

On Tue, 8 Feb 2022 13:40:14 GMT, Roland Westrelin <roland at openjdk.org> wrote:

> I think this is caused by the support that was added recently for
> virtual call late inlining and I propose tweaking the
> AlwaysIncrementalInline logic for that case.

This pull request has now been integrated.

Changeset: 18b1f28d
Author:    Roland Westrelin <roland at openjdk.org>
URL:       https://git.openjdk.java.net/valhalla/commit/18b1f28d13eeb7c390fd9298be579801799562a3
Stats:     3 lines in 1 file changed: 1 ins; 0 del; 2 mod

8281378: [lworld] Crash in LateInlineMHCallGenerator::do_late_inline_check

Reviewed-by: thartmann

-------------

PR: https://git.openjdk.java.net/valhalla/pull/637

From fparain at openjdk.java.net  Tue Feb  8 15:12:36 2022
From: fparain at openjdk.java.net (Frederic Parain)
Date: Tue, 8 Feb 2022 15:12:36 GMT
Subject: [lworld] RFR: 8281279: [lworld] Add JVM support for
 ACC_PERMITS_VALUE [v5]
In-Reply-To: <txivvMsAReynqMQjpKRvJffLaiEjfi1ZgCfj9zZy-7Y=.1afa863d-f2c3-47f2-a9e2-d916e3beaefb@github.com>
References: <Bx7Aqv2sFJ4TMvasfAyOIfxdc661Pm23M1GCzy6nFGQ=.3abe36fd-1b91-4ac9-82ed-e67ad16d74f7@github.com>
 <txivvMsAReynqMQjpKRvJffLaiEjfi1ZgCfj9zZy-7Y=.1afa863d-f2c3-47f2-a9e2-d916e3beaefb@github.com>
Message-ID: <Yhm3RvHh-c6zcJjwxEhC1P2pVia_IptsK2gCqJlcRaQ=.d6af5a14-86d4-4bac-82be-c535ef5c7a26@github.com>

On Mon, 7 Feb 2022 18:15:54 GMT, Harold Seigel <hseigel at openjdk.org> wrote:

>> Please review this fix to add JVM support for ACC_PERMITS_VALUE.  This change throws an ICCE exception when loading a value class whose super class is not java.lang.Object or does not have ACC_PERMITS_VALUE set.  The fix was tested with Mach5 tiers 1-2 on Linux, Mac OS, and Windows and Mach5 tiers 3-5 on Linux x64.
>> 
>> Thanks, Harold
>
> Harold Seigel has updated the pull request incrementally with one additional commit since the last revision:
> 
>   add j.l.Object comment

Marked as reviewed by fparain (Committer).

-------------

PR: https://git.openjdk.java.net/valhalla/pull/632

From hseigel at openjdk.java.net  Tue Feb  8 15:24:23 2022
From: hseigel at openjdk.java.net (Harold Seigel)
Date: Tue, 8 Feb 2022 15:24:23 GMT
Subject: [lworld] RFR: 8281279: [lworld] Add JVM support for
 ACC_PERMITS_VALUE [v5]
In-Reply-To: <txivvMsAReynqMQjpKRvJffLaiEjfi1ZgCfj9zZy-7Y=.1afa863d-f2c3-47f2-a9e2-d916e3beaefb@github.com>
References: <Bx7Aqv2sFJ4TMvasfAyOIfxdc661Pm23M1GCzy6nFGQ=.3abe36fd-1b91-4ac9-82ed-e67ad16d74f7@github.com>
 <txivvMsAReynqMQjpKRvJffLaiEjfi1ZgCfj9zZy-7Y=.1afa863d-f2c3-47f2-a9e2-d916e3beaefb@github.com>
Message-ID: <HAmZ2cQn0bx2s8HF8A9p56Ge5j52NIPCieMM0_gLvng=.328cab03-3087-4695-ae21-e60546248687@github.com>

On Mon, 7 Feb 2022 18:15:54 GMT, Harold Seigel <hseigel at openjdk.org> wrote:

>> Please review this fix to add JVM support for ACC_PERMITS_VALUE.  This change throws an ICCE exception when loading a value class whose super class is not java.lang.Object or does not have ACC_PERMITS_VALUE set.  The fix was tested with Mach5 tiers 1-2 on Linux, Mac OS, and Windows and Mach5 tiers 3-5 on Linux x64.
>> 
>> Thanks, Harold
>
> Harold Seigel has updated the pull request incrementally with one additional commit since the last revision:
> 
>   add j.l.Object comment

Thanks Fred for the review!

-------------

PR: https://git.openjdk.java.net/valhalla/pull/632

From hseigel at openjdk.java.net  Tue Feb  8 15:24:23 2022
From: hseigel at openjdk.java.net (Harold Seigel)
Date: Tue, 8 Feb 2022 15:24:23 GMT
Subject: [lworld] Integrated: 8281279: [lworld] Add JVM support for
 ACC_PERMITS_VALUE
In-Reply-To: <Bx7Aqv2sFJ4TMvasfAyOIfxdc661Pm23M1GCzy6nFGQ=.3abe36fd-1b91-4ac9-82ed-e67ad16d74f7@github.com>
References: <Bx7Aqv2sFJ4TMvasfAyOIfxdc661Pm23M1GCzy6nFGQ=.3abe36fd-1b91-4ac9-82ed-e67ad16d74f7@github.com>
Message-ID: <yOP-SI4o4nKpJ_hyNJ3LzIk-n4QvBNkKyV_9BQiBU8g=.448f9424-d731-4fc8-94c0-2de3ff37d1ef@github.com>

On Mon, 7 Feb 2022 13:34:47 GMT, Harold Seigel <hseigel at openjdk.org> wrote:

> Please review this fix to add JVM support for ACC_PERMITS_VALUE.  This change throws an ICCE exception when loading a value class whose super class is not java.lang.Object or does not have ACC_PERMITS_VALUE set.  The fix was tested with Mach5 tiers 1-2 on Linux, Mac OS, and Windows and Mach5 tiers 3-5 on Linux x64.
> 
> Thanks, Harold

This pull request has now been integrated.

Changeset: 3017cf71
Author:    Harold Seigel <hseigel at openjdk.org>
URL:       https://git.openjdk.java.net/valhalla/commit/3017cf71c2201b236415c671827cfc3bac9c3521
Stats:     1668 lines in 17 files changed: 1367 ins; 205 del; 96 mod

8281279: [lworld] Add JVM support for ACC_PERMITS_VALUE

Reviewed-by: fparain

-------------

PR: https://git.openjdk.java.net/valhalla/pull/632

From fparain at openjdk.java.net  Tue Feb  8 15:56:51 2022
From: fparain at openjdk.java.net (Frederic Parain)
Date: Tue, 8 Feb 2022 15:56:51 GMT
Subject: [lworld] RFR: 8281380: [lworld] Rename T_INLINE_TYPE to
 T_PRIMITIVE_OBJECT
Message-ID: <MPGHNsV012KLyY9JfD_q2KrrlyC4rKRtEGjfUSxrn20=.90bdbb91-f9dc-4d72-bce2-81d7e949a857@github.com>

Please review this renaming patch, changing T_INLINE_TYPE to T_PRIMITIVE_OBJECT in order to prevent the confusion between the InlineKlass concept, which is used to represent both value classes and primitive classes, and the BasicType that represents only primitive classes.

Remaining occurrences of the "INLINE_TYPE" string are located in a JDI test, which could be fixed later, and in the verifier code, which might need to be revisited to check the way value classes/primitive classes are handled.

Tested with Mach5, tier1 to ensure build is working on all platforms.

Thank you,

Fred

-------------

Commit messages:
 - Rename Basictype T_INLINE_TYPE to T_PRIMITIVE_OBJECT

Changes: https://git.openjdk.java.net/valhalla/pull/638/files
 Webrev: https://webrevs.openjdk.java.net/?repo=valhalla&pr=638&range=00
  Issue: https://bugs.openjdk.java.net/browse/JDK-8281380
  Stats: 323 lines in 98 files changed: 0 ins; 0 del; 323 mod
  Patch: https://git.openjdk.java.net/valhalla/pull/638.diff
  Fetch: git fetch https://git.openjdk.java.net/valhalla pull/638/head:pull/638

PR: https://git.openjdk.java.net/valhalla/pull/638

From hseigel at openjdk.java.net  Tue Feb  8 16:18:25 2022
From: hseigel at openjdk.java.net (Harold Seigel)
Date: Tue, 8 Feb 2022 16:18:25 GMT
Subject: [lworld] RFR: 8281380: [lworld] Rename T_INLINE_TYPE to
 T_PRIMITIVE_OBJECT
In-Reply-To: <MPGHNsV012KLyY9JfD_q2KrrlyC4rKRtEGjfUSxrn20=.90bdbb91-f9dc-4d72-bce2-81d7e949a857@github.com>
References: <MPGHNsV012KLyY9JfD_q2KrrlyC4rKRtEGjfUSxrn20=.90bdbb91-f9dc-4d72-bce2-81d7e949a857@github.com>
Message-ID: <JMMgxa1Nm3uZeO16Vk2lBNEad7xHfa1AS_GA9ohsGRw=.f900d19d-9f9c-471b-86e5-c66eb8c39113@github.com>

On Tue, 8 Feb 2022 15:51:34 GMT, Frederic Parain <fparain at openjdk.org> wrote:

> Please review this renaming patch, changing T_INLINE_TYPE to T_PRIMITIVE_OBJECT in order to prevent the confusion between the InlineKlass concept, which is used to represent both value classes and primitive classes, and the BasicType that represents only primitive classes.
> 
> Remaining occurrences of the "INLINE_TYPE" string are located in a JDI test, which could be fixed later, and in the verifier code, which might need to be revisited to check the way value classes/primitive classes are handled.
> 
> Tested with Mach5, tier1 to ensure build is working on all platforms.
> 
> Thank you,
> 
> Fred

Changes look good!  I wonder if src/java.base/share/native/libverify/check_code.c needs to worry about T_PRIMITVE_OBJECT's.  It should only see old classfiles.

-------------

Marked as reviewed by hseigel (Committer).

PR: https://git.openjdk.java.net/valhalla/pull/638

From vromero at openjdk.java.net  Tue Feb  8 18:02:47 2022
From: vromero at openjdk.java.net (Vicente Romero)
Date: Tue, 8 Feb 2022 18:02:47 GMT
Subject: RFR: some existing regression tests are failing due to universal tvars
 changes
Message-ID: <FWQxROE--O9V03_n8iVqy-26wAlK-mFGZPAc6sn8sJk=.59080022-146e-4d35-a3c0-e579737840a3@github.com>

This PR contains some type system bug fixes that were impacting existing regression tests. Improved code at Attr that deals with the `universality` of declared type variables in order to consider if primitives could be accepted as bounds in wildcards etc

-------------

Commit messages:
 - some existing regression tests are failing due to universal tvars changes

Changes: https://git.openjdk.java.net/valhalla/pull/639/files
 Webrev: https://webrevs.openjdk.java.net/?repo=valhalla&pr=639&range=00
  Stats: 68 lines in 4 files changed: 53 ins; 4 del; 11 mod
  Patch: https://git.openjdk.java.net/valhalla/pull/639.diff
  Fetch: git fetch https://git.openjdk.java.net/valhalla pull/639/head:pull/639

PR: https://git.openjdk.java.net/valhalla/pull/639

From vromero at openjdk.java.net  Tue Feb  8 18:13:34 2022
From: vromero at openjdk.java.net (Vicente Romero)
Date: Tue, 8 Feb 2022 18:13:34 GMT
Subject: Integrated: some existing regression tests are failing due to
 universal tvars changes
In-Reply-To: <FWQxROE--O9V03_n8iVqy-26wAlK-mFGZPAc6sn8sJk=.59080022-146e-4d35-a3c0-e579737840a3@github.com>
References: <FWQxROE--O9V03_n8iVqy-26wAlK-mFGZPAc6sn8sJk=.59080022-146e-4d35-a3c0-e579737840a3@github.com>
Message-ID: <J5bEC4XYQtvZaI0ujfWKx4vzFUQT7rtr2OsDPx1fxO4=.c970099a-e146-4632-ba50-0d3bd920a49a@github.com>

On Tue, 8 Feb 2022 17:17:45 GMT, Vicente Romero <vromero at openjdk.org> wrote:

> This PR contains some type system bug fixes that were impacting existing regression tests. Improved code at Attr that deals with the `universality` of declared type variables in order to consider if primitives could be accepted as bounds in wildcards etc

This pull request has now been integrated.

Changeset: 79d60ae4
Author:    Vicente Romero <vromero at openjdk.org>
URL:       https://git.openjdk.java.net/valhalla/commit/79d60ae4e10c32e457acf9b7e961c0b3bd197cd8
Stats:     68 lines in 4 files changed: 53 ins; 4 del; 11 mod

some existing regression tests are failing due to universal tvars changes

-------------

PR: https://git.openjdk.java.net/valhalla/pull/639

From fparain at openjdk.java.net  Tue Feb  8 18:54:29 2022
From: fparain at openjdk.java.net (Frederic Parain)
Date: Tue, 8 Feb 2022 18:54:29 GMT
Subject: [lworld] RFR: 8281380: [lworld] Rename T_INLINE_TYPE to
 T_PRIMITIVE_OBJECT
In-Reply-To: <JMMgxa1Nm3uZeO16Vk2lBNEad7xHfa1AS_GA9ohsGRw=.f900d19d-9f9c-471b-86e5-c66eb8c39113@github.com>
References: <MPGHNsV012KLyY9JfD_q2KrrlyC4rKRtEGjfUSxrn20=.90bdbb91-f9dc-4d72-bce2-81d7e949a857@github.com>
 <JMMgxa1Nm3uZeO16Vk2lBNEad7xHfa1AS_GA9ohsGRw=.f900d19d-9f9c-471b-86e5-c66eb8c39113@github.com>
Message-ID: <NswKhsQ2rUb6V-ecacKY0d4a3jv0N2Uoit7pwZXvf-0=.1b2966dc-7859-4534-95bb-fd95013b7728@github.com>

On Tue, 8 Feb 2022 16:14:56 GMT, Harold Seigel <hseigel at openjdk.org> wrote:

>  I wonder if src/java.base/share/native/libverify/check_code.c needs to worry about T_PRIMITVE_OBJECT's. It should only see old classfiles.

check_code.c might not need to support Q-signatures, but the additional logic was added back in 2018.
I'd prefer to limit this changeset to pure renaming, and handle the case of check_code.c in a different changeset.

-------------

PR: https://git.openjdk.java.net/valhalla/pull/638

From fparain at openjdk.java.net  Tue Feb  8 19:51:35 2022
From: fparain at openjdk.java.net (Frederic Parain)
Date: Tue, 8 Feb 2022 19:51:35 GMT
Subject: [lworld] Integrated: 8281380: [lworld] Rename T_INLINE_TYPE to
 T_PRIMITIVE_OBJECT
In-Reply-To: <MPGHNsV012KLyY9JfD_q2KrrlyC4rKRtEGjfUSxrn20=.90bdbb91-f9dc-4d72-bce2-81d7e949a857@github.com>
References: <MPGHNsV012KLyY9JfD_q2KrrlyC4rKRtEGjfUSxrn20=.90bdbb91-f9dc-4d72-bce2-81d7e949a857@github.com>
Message-ID: <XDhtASf7hEV9xB80IpgdXq1EHZFX5crOys_fXlPRQxc=.d847812d-b05c-4b22-8c4b-5523bd679c75@github.com>

On Tue, 8 Feb 2022 15:51:34 GMT, Frederic Parain <fparain at openjdk.org> wrote:

> Please review this renaming patch, changing T_INLINE_TYPE to T_PRIMITIVE_OBJECT in order to prevent the confusion between the InlineKlass concept, which is used to represent both value classes and primitive classes, and the BasicType that represents only primitive classes.
> 
> Remaining occurrences of the "INLINE_TYPE" string are located in a JDI test, which could be fixed later, and in the verifier code, which might need to be revisited to check the way value classes/primitive classes are handled.
> 
> Tested with Mach5, tier1 to ensure build is working on all platforms.
> 
> Thank you,
> 
> Fred

This pull request has now been integrated.

Changeset: 954f5ca4
Author:    Frederic Parain <fparain at openjdk.org>
URL:       https://git.openjdk.java.net/valhalla/commit/954f5ca407e4d68204fc0010d2082103255a61cb
Stats:     323 lines in 98 files changed: 0 ins; 0 del; 323 mod

8281380: [lworld] Rename T_INLINE_TYPE to T_PRIMITIVE_OBJECT

Reviewed-by: hseigel

-------------

PR: https://git.openjdk.java.net/valhalla/pull/638

From rriggs at openjdk.java.net  Tue Feb  8 21:29:56 2022
From: rriggs at openjdk.java.net (Roger Riggs)
Date: Tue, 8 Feb 2022 21:29:56 GMT
Subject: [lworld] RFR: 8281463: [lworld] VALUE / PRIMITIVE modifiers should be
 supported by reflection
Message-ID: <H11q_a517sX2Ta6V9p9meFxazzPRhYO7xtKon6DhrVQ=.17fc9f8d-05ad-4b32-ae82-30f4650ee13b@github.com>

Update java.lang.reflection.Modifier to reflect the JVMS specified bits in the modifier.

The re-use of bits between method and class modifiers poses a problem for this API.
Previously, all bits were distinct and the modifiers do not include any indication of whether
the modifier bits are from a class, field, or method. As a result there is ambiguity as two which
of two meanings a bit takes on.  NATIVE vs VALUE and STRICT vs PRIMITIVE.

Please review for discussion purposes.

-------------

Commit messages:
 - 8281463: [lworld] VALUE / PRIMITIVE modifiers should be supported by reflection

Changes: https://git.openjdk.java.net/valhalla/pull/640/files
 Webrev: https://webrevs.openjdk.java.net/?repo=valhalla&pr=640&range=00
  Issue: https://bugs.openjdk.java.net/browse/JDK-8281463
  Stats: 76 lines in 3 files changed: 44 ins; 8 del; 24 mod
  Patch: https://git.openjdk.java.net/valhalla/pull/640.diff
  Fetch: git fetch https://git.openjdk.java.net/valhalla pull/640/head:pull/640

PR: https://git.openjdk.java.net/valhalla/pull/640

From vromero at openjdk.java.net  Tue Feb  8 22:13:36 2022
From: vromero at openjdk.java.net (Vicente Romero)
Date: Tue, 8 Feb 2022 22:13:36 GMT
Subject: Integrated: Merge lworld
Message-ID: <ZXNizzOBX5ltksSLKESd9a7FNw_0vBAVT5VczieT4zo=.d9b8c07b-d2a7-446a-aebd-3ff4b9cbda18@github.com>

Merge lworld into universal-tvars

# Conflicts:
#	src/jdk.compiler/share/classes/com/sun/tools/javac/code/Flags.java
#	src/jdk.compiler/share/classes/com/sun/tools/javac/code/Source.java
#	src/jdk.compiler/share/classes/com/sun/tools/javac/code/Types.java
#	src/jdk.compiler/share/classes/com/sun/tools/javac/comp/Attr.java
#	src/jdk.compiler/share/classes/com/sun/tools/javac/parser/JavacParser.java
#	src/jdk.compiler/share/classes/com/sun/tools/javac/resources/compiler.properties
#	test/langtools/tools/javac/diags/examples.not-yet.txt

-------------

Commit messages:
 - Merge branch 'lworld' into universal-tvars_merge_lworld
 - some existing regression tests are failing due to universal tvars changes
 - Adding type system regression tests
 - some bug fixing, adding new regression tests
 - Merge lworld
 - Primitive value conversion
 - updating tests
 - Merge lworld
 - Merge lworld
 - universal type variables: initial prototype
 - ... and 3 more: https://git.openjdk.java.net/valhalla/compare/954f5ca4...b249ca68

The merge commit only contains trivial merges, so no merge-specific webrevs have been generated.

Changes: https://git.openjdk.java.net/valhalla/pull/641/files
  Stats: 1187 lines in 30 files changed: 1081 ins; 26 del; 80 mod
  Patch: https://git.openjdk.java.net/valhalla/pull/641.diff
  Fetch: git fetch https://git.openjdk.java.net/valhalla pull/641/head:pull/641

PR: https://git.openjdk.java.net/valhalla/pull/641

From vromero at openjdk.java.net  Tue Feb  8 22:13:39 2022
From: vromero at openjdk.java.net (Vicente Romero)
Date: Tue, 8 Feb 2022 22:13:39 GMT
Subject: Integrated: Merge lworld
In-Reply-To: <ZXNizzOBX5ltksSLKESd9a7FNw_0vBAVT5VczieT4zo=.d9b8c07b-d2a7-446a-aebd-3ff4b9cbda18@github.com>
References: <ZXNizzOBX5ltksSLKESd9a7FNw_0vBAVT5VczieT4zo=.d9b8c07b-d2a7-446a-aebd-3ff4b9cbda18@github.com>
Message-ID: <YZh9ajRbfVwMvFLxADJwkwm8Xkl-B9VBMGsNuHhEq-A=.dcfb984b-5461-412b-bbc8-342811630c22@github.com>

On Tue, 8 Feb 2022 21:59:07 GMT, Vicente Romero <vromero at openjdk.org> wrote:

> Merge lworld into universal-tvars
> 
> # Conflicts:
> #	src/jdk.compiler/share/classes/com/sun/tools/javac/code/Flags.java
> #	src/jdk.compiler/share/classes/com/sun/tools/javac/code/Source.java
> #	src/jdk.compiler/share/classes/com/sun/tools/javac/code/Types.java
> #	src/jdk.compiler/share/classes/com/sun/tools/javac/comp/Attr.java
> #	src/jdk.compiler/share/classes/com/sun/tools/javac/parser/JavacParser.java
> #	src/jdk.compiler/share/classes/com/sun/tools/javac/resources/compiler.properties
> #	test/langtools/tools/javac/diags/examples.not-yet.txt

This pull request has now been integrated.

Changeset: c1a11fc6
Author:    Vicente Romero <vromero at openjdk.org>
URL:       https://git.openjdk.java.net/valhalla/commit/c1a11fc60bd4753ecd74cf7abbaade3cf98bbb50
Stats:     111302 lines in 3055 files changed: 73584 ins; 21472 del; 16246 mod

Merge lworld

-------------

PR: https://git.openjdk.java.net/valhalla/pull/641

From mchung at openjdk.java.net  Tue Feb  8 22:36:33 2022
From: mchung at openjdk.java.net (Mandy Chung)
Date: Tue, 8 Feb 2022 22:36:33 GMT
Subject: [lworld] RFR: 8281463: [lworld] VALUE / PRIMITIVE modifiers should
 be supported by reflection
In-Reply-To: <H11q_a517sX2Ta6V9p9meFxazzPRhYO7xtKon6DhrVQ=.17fc9f8d-05ad-4b32-ae82-30f4650ee13b@github.com>
References: <H11q_a517sX2Ta6V9p9meFxazzPRhYO7xtKon6DhrVQ=.17fc9f8d-05ad-4b32-ae82-30f4650ee13b@github.com>
Message-ID: <32sUMw_FT6ytRNJhSwpty5xN8VPYaEOlNABCkdsLyxE=.06277f7c-413c-4144-aea2-3a70e5f8c741@github.com>

On Tue, 8 Feb 2022 21:23:30 GMT, Roger Riggs <rriggs at openjdk.org> wrote:

> Update java.lang.reflection.Modifier to reflect the JVMS specified bits in the modifier.
> 
> The re-use of bits between method and class modifiers poses a problem for this API.
> Previously, all bits were distinct and the modifiers do not include any indication of whether
> the modifier bits are from a class, field, or method. As a result there is ambiguity as two which
> of two meanings a bit takes on.  NATIVE vs VALUE and STRICT vs PRIMITIVE.
> 
> Please review for discussion purposes.

`ACC_VALUE`, `ACC_PRIMITIVE`, and `ACC_PERMITS_VALUE` are class modifiers.   Not every class modifiers defined in JVMS  is defined in the `java.lang.reflect.Modifier` API for example ACC_ENUM, ACC_ANNOTATION and ACC_MODULE.   But instead a test method such as `Class::isEnum`, `Class::isAnnotation` is provided in the `java.lang.Class` API.

For value and primitive class, a new `Class::isValue` and `Class::isPrimitiveClass` API (subject to the final terminology for B2 and B3) would suffice.   I'm not sure yet if an API is needed for `ACC_PERMITS_VALUE` since this is mostly used by JVM.

-------------

PR: https://git.openjdk.java.net/valhalla/pull/640

From duke at openjdk.java.net  Wed Feb  9 10:59:39 2022
From: duke at openjdk.java.net (Aggelos Biboudis)
Date: Wed, 9 Feb 2022 10:59:39 GMT
Subject: [lworld] RFR: 8281026: Allow for
 compiler.note.cant.instantiate.object.directly to be suppressed via an option
In-Reply-To: <76skw4CeVr0i6WFPBK4ToC3eQc661eq_0QKi8TjKob0=.ea4ad890-0e7f-4357-a4fb-dc85cb0de50e@github.com>
References: <76skw4CeVr0i6WFPBK4ToC3eQc661eq_0QKi8TjKob0=.ea4ad890-0e7f-4357-a4fb-dc85cb0de50e@github.com>
Message-ID: <J4lt79d5RIzcd-xRTPQke4Pr13PHDw-lXsbs1VL_o7c=.169d0aa2-37db-45d7-878e-3958a948435f@github.com>

On Tue, 8 Feb 2022 12:44:33 GMT, Srikanth Adayapalam <sadayapalam at openjdk.org> wrote:

> Introduce the javac option -XDtolerateObjectInstantiation to suppress the warning.

It's less verbose now. ??

Marked as reviewed by biboudis at github.com (no known OpenJDK username).

-------------

PR: https://git.openjdk.java.net/valhalla/pull/636

From thartmann at openjdk.java.net  Wed Feb  9 11:11:35 2022
From: thartmann at openjdk.java.net (Tobias Hartmann)
Date: Wed, 9 Feb 2022 11:11:35 GMT
Subject: [lworld] RFR: 8281454: [lworld] Assert in EA due to oop access to flat
 array
Message-ID: <J2mdNVWRr_gLMl4WStGICazyXL5RnZHoYviMwGC6Qi0=.971420ef-98ae-490b-82a5-a55a96e299d9@github.com>

Similar to [JDK-8250951](https://bugs.openjdk.java.net/browse/JDK-8250951), we hit an assert in EA because an oop load is used to access a flat array. The load is from an `Arrays.copyOf` and flat array checks in `LibraryCallKit::inline_array_copyOf` will prevent that load from ever being executed but the dead subgraph is not removed because we run with `-XX:+StressReflectiveCode`.

I fixed this by using the `FlatArrayCheckNode` for checks in the intrinsics as well. This ensures that the graph is folded even with  `-XX:+StressReflectiveCode` and enables more aggressive optimizations. The patch also includes some cleanups and removal of (now) dead code.

Thanks,
Tobias

-------------

Commit messages:
 - Removed override keywords
 - 8281454: [lworld] Assert in EA due to oop access to flat array

Changes: https://git.openjdk.java.net/valhalla/pull/642/files
 Webrev: https://webrevs.openjdk.java.net/?repo=valhalla&pr=642&range=00
  Issue: https://bugs.openjdk.java.net/browse/JDK-8281454
  Stats: 83 lines in 12 files changed: 10 ins; 22 del; 51 mod
  Patch: https://git.openjdk.java.net/valhalla/pull/642.diff
  Fetch: git fetch https://git.openjdk.java.net/valhalla pull/642/head:pull/642

PR: https://git.openjdk.java.net/valhalla/pull/642

From thartmann at openjdk.java.net  Wed Feb  9 11:39:37 2022
From: thartmann at openjdk.java.net (Tobias Hartmann)
Date: Wed, 9 Feb 2022 11:39:37 GMT
Subject: [lworld] Integrated: 8281454: [lworld] Assert in EA due to oop access
 to flat array
In-Reply-To: <J2mdNVWRr_gLMl4WStGICazyXL5RnZHoYviMwGC6Qi0=.971420ef-98ae-490b-82a5-a55a96e299d9@github.com>
References: <J2mdNVWRr_gLMl4WStGICazyXL5RnZHoYviMwGC6Qi0=.971420ef-98ae-490b-82a5-a55a96e299d9@github.com>
Message-ID: <t6y4AmrOjUfpnven4Wl7oKunjJJUsIJOUTNYiqI95lc=.d647fbcd-1180-4fcc-9936-0f08aeb2be2d@github.com>

On Wed, 9 Feb 2022 11:05:10 GMT, Tobias Hartmann <thartmann at openjdk.org> wrote:

> Similar to [JDK-8250951](https://bugs.openjdk.java.net/browse/JDK-8250951), we hit an assert in EA because an oop load is used to access a flat array. The load is from an `Arrays.copyOf` and flat array checks in `LibraryCallKit::inline_array_copyOf` will prevent that load from ever being executed but the dead subgraph is not removed because we run with `-XX:+StressReflectiveCode`.
> 
> I fixed this by using the `FlatArrayCheckNode` for checks in the intrinsics as well. This ensures that the graph is folded even with  `-XX:+StressReflectiveCode` and enables more aggressive optimizations. The patch also includes some cleanups and removal of (now) dead code.
> 
> Thanks,
> Tobias

This pull request has now been integrated.

Changeset: 5a0f3336
Author:    Tobias Hartmann <thartmann at openjdk.org>
URL:       https://git.openjdk.java.net/valhalla/commit/5a0f3336c0e51ea8e4942f7b70a40f37ded0c2f2
Stats:     83 lines in 12 files changed: 10 ins; 22 del; 51 mod

8281454: [lworld] Assert in EA due to oop access to flat array

-------------

PR: https://git.openjdk.java.net/valhalla/pull/642

From sadayapalam at openjdk.java.net  Wed Feb  9 12:29:34 2022
From: sadayapalam at openjdk.java.net (Srikanth Adayapalam)
Date: Wed, 9 Feb 2022 12:29:34 GMT
Subject: [lworld] RFR: 8281026: Allow for
 compiler.note.cant.instantiate.object.directly to be suppressed via an option
In-Reply-To: <76skw4CeVr0i6WFPBK4ToC3eQc661eq_0QKi8TjKob0=.ea4ad890-0e7f-4357-a4fb-dc85cb0de50e@github.com>
References: <76skw4CeVr0i6WFPBK4ToC3eQc661eq_0QKi8TjKob0=.ea4ad890-0e7f-4357-a4fb-dc85cb0de50e@github.com>
Message-ID: <NhSVKAt13kgOclQuEjJ_L2pnBbNtwe52eXKVPgpOLqM=.cd43d4a0-01dd-47cf-ac45-3898e342c049@github.com>

On Tue, 8 Feb 2022 12:44:33 GMT, Srikanth Adayapalam <sadayapalam at openjdk.org> wrote:

> Introduce the javac option -XDtolerateObjectInstantiation to suppress the warning.

Thanks for the review Aggelos!

-------------

PR: https://git.openjdk.java.net/valhalla/pull/636

From sadayapalam at openjdk.java.net  Wed Feb  9 12:29:35 2022
From: sadayapalam at openjdk.java.net (Srikanth Adayapalam)
Date: Wed, 9 Feb 2022 12:29:35 GMT
Subject: [lworld] Integrated: 8281026: Allow for
 compiler.note.cant.instantiate.object.directly to be suppressed via an option
In-Reply-To: <76skw4CeVr0i6WFPBK4ToC3eQc661eq_0QKi8TjKob0=.ea4ad890-0e7f-4357-a4fb-dc85cb0de50e@github.com>
References: <76skw4CeVr0i6WFPBK4ToC3eQc661eq_0QKi8TjKob0=.ea4ad890-0e7f-4357-a4fb-dc85cb0de50e@github.com>
Message-ID: <1nhx0NcjoMWru2cYkXpTjhORECqbu6YgPCPmoS8wDys=.1267c4cc-020b-4358-b8ab-0cff6b663f6b@github.com>

On Tue, 8 Feb 2022 12:44:33 GMT, Srikanth Adayapalam <sadayapalam at openjdk.org> wrote:

> Introduce the javac option -XDtolerateObjectInstantiation to suppress the warning.

This pull request has now been integrated.

Changeset: fe2126b2
Author:    Srikanth Adayapalam <sadayapalam at openjdk.org>
URL:       https://git.openjdk.java.net/valhalla/commit/fe2126b2fc2c865fc2d1bbc7bb8154fa00c678da
Stats:     20 lines in 4 files changed: 14 ins; 0 del; 6 mod

8281026: Allow for compiler.note.cant.instantiate.object.directly to be suppressed via an option

-------------

PR: https://git.openjdk.java.net/valhalla/pull/636

From fparain at openjdk.java.net  Wed Feb  9 13:34:09 2022
From: fparain at openjdk.java.net (Frederic Parain)
Date: Wed, 9 Feb 2022 13:34:09 GMT
Subject: [lworld] Integrated: 8281537: [lworld]
 runtime/cds/appcds/TestDumpClassListSource.java fails with Exception:
 java.lang.RuntimeException: Class Hello should be printed in classlist
Message-ID: <i9aQmYN1NT3hMCUoB12HBmPBhDQ3Qkdk2Uyw9jwskRE=.da163ee8-c261-44ba-a859-d02f9663f361@github.com>

Small test fix.

-------------

Commit messages:
 - Fix pattern matching to expect injected interface

Changes: https://git.openjdk.java.net/valhalla/pull/643/files
 Webrev: https://webrevs.openjdk.java.net/?repo=valhalla&pr=643&range=00
  Issue: https://bugs.openjdk.java.net/browse/JDK-8281537
  Stats: 1 line in 1 file changed: 0 ins; 0 del; 1 mod
  Patch: https://git.openjdk.java.net/valhalla/pull/643.diff
  Fetch: git fetch https://git.openjdk.java.net/valhalla pull/643/head:pull/643

PR: https://git.openjdk.java.net/valhalla/pull/643

From fparain at openjdk.java.net  Wed Feb  9 13:34:10 2022
From: fparain at openjdk.java.net (Frederic Parain)
Date: Wed, 9 Feb 2022 13:34:10 GMT
Subject: [lworld] Integrated: 8281537: [lworld]
 runtime/cds/appcds/TestDumpClassListSource.java fails with Exception:
 java.lang.RuntimeException: Class Hello should be printed in classlist
In-Reply-To: <i9aQmYN1NT3hMCUoB12HBmPBhDQ3Qkdk2Uyw9jwskRE=.da163ee8-c261-44ba-a859-d02f9663f361@github.com>
References: <i9aQmYN1NT3hMCUoB12HBmPBhDQ3Qkdk2Uyw9jwskRE=.da163ee8-c261-44ba-a859-d02f9663f361@github.com>
Message-ID: <OrcUEWCWYCV-7GD1P9HE9cgoLrn2c-fs9edilgT8ZKw=.48634489-6d02-4c58-8a15-bbcfae7c3fae@github.com>

On Wed, 9 Feb 2022 13:27:49 GMT, Frederic Parain <fparain at openjdk.org> wrote:

> Small test fix.

This pull request has now been integrated.

Changeset: 9cab68b4
Author:    Frederic Parain <fparain at openjdk.org>
URL:       https://git.openjdk.java.net/valhalla/commit/9cab68b494b7d89f7782a50cb67e7677e7253729
Stats:     1 line in 1 file changed: 0 ins; 0 del; 1 mod

8281537: [lworld] runtime/cds/appcds/TestDumpClassListSource.java fails with Exception: java.lang.RuntimeException: Class Hello should be printed in classlist

-------------

PR: https://git.openjdk.java.net/valhalla/pull/643

From mchung at openjdk.java.net  Wed Feb  9 20:31:11 2022
From: mchung at openjdk.java.net (Mandy Chung)
Date: Wed, 9 Feb 2022 20:31:11 GMT
Subject: [lworld] RFR: 8281073: [lworld] Update MethodHandles::zero, empty,
 constant APIs for primitive classes
Message-ID: <XgZFb4_E4GhjfuXo8DPw6bmynBOJMDmGIqwBKMR2vYQ=.b828f472-c86f-427e-8d42-1fc198c87899@github.com>

The `MethodHandles::zero` and `MethodHandles::empty` APIs are updated to return a primitive object with the default value if the given type is a primitive value type. These methods return null if the given type is a value class and a primitive reference type.

The MethodHandles::constant method will throw `NullPointerException` if the given type is a primitive value type and the given value is null.   That's the same behavior as the `int` type but such case is missing in the spec.

-------------

Commit messages:
 - Further clean up
 - JDK-8281073: [lworld] Update MethodHandles::zero, empty, constant APIs for value classes

Changes: https://git.openjdk.java.net/valhalla/pull/644/files
 Webrev: https://webrevs.openjdk.java.net/?repo=valhalla&pr=644&range=00
  Issue: https://bugs.openjdk.java.net/browse/JDK-8281073
  Stats: 132 lines in 2 files changed: 126 ins; 0 del; 6 mod
  Patch: https://git.openjdk.java.net/valhalla/pull/644.diff
  Fetch: git fetch https://git.openjdk.java.net/valhalla pull/644/head:pull/644

PR: https://git.openjdk.java.net/valhalla/pull/644

From amenkov at openjdk.java.net  Thu Feb 10 15:09:48 2022
From: amenkov at openjdk.java.net (Alex Menkov)
Date: Thu, 10 Feb 2022 15:09:48 GMT
Subject: [lworld] RFR: 8281073: [lworld] Update MethodHandles::zero, empty,
 constant APIs for primitive classes
In-Reply-To: <XgZFb4_E4GhjfuXo8DPw6bmynBOJMDmGIqwBKMR2vYQ=.b828f472-c86f-427e-8d42-1fc198c87899@github.com>
References: <XgZFb4_E4GhjfuXo8DPw6bmynBOJMDmGIqwBKMR2vYQ=.b828f472-c86f-427e-8d42-1fc198c87899@github.com>
Message-ID: <RxFmJ7iy_C-RNqj5YCYe5xBLAUs42t_A4WxoFe7TvpQ=.7a7ca52b-5b2f-4395-8065-7d8c0f19055b@github.com>

On Wed, 9 Feb 2022 20:24:07 GMT, Mandy Chung <mchung at openjdk.org> wrote:

> The `MethodHandles::zero` and `MethodHandles::empty` APIs are updated to return a primitive object with the default value if the given type is a primitive value type. These methods return null if the given type is a value class and a primitive reference type.
> 
> The MethodHandles::constant method will throw `NullPointerException` if the given type is a primitive value type and the given value is null.   That's the same behavior as the `int` type but such case is missing in the spec.

src/java.base/share/classes/java/lang/invoke/MethodHandles.java line 5116:

> 5114:      * and returns a suitable default depending on the return type.
> 5115:      * If the requested type is a primitive type or {@code void}, it returns
> 5116:      * a zero primitive value or {@ocde void}.

{@ocde void} -> {@code void}

-------------

PR: https://git.openjdk.java.net/valhalla/pull/644

From mchung at openjdk.java.net  Thu Feb 10 18:13:25 2022
From: mchung at openjdk.java.net (Mandy Chung)
Date: Thu, 10 Feb 2022 18:13:25 GMT
Subject: [lworld] RFR: 8281073: [lworld] Update MethodHandles::zero, empty,
 constant APIs for primitive classes [v2]
In-Reply-To: <XgZFb4_E4GhjfuXo8DPw6bmynBOJMDmGIqwBKMR2vYQ=.b828f472-c86f-427e-8d42-1fc198c87899@github.com>
References: <XgZFb4_E4GhjfuXo8DPw6bmynBOJMDmGIqwBKMR2vYQ=.b828f472-c86f-427e-8d42-1fc198c87899@github.com>
Message-ID: <ggKUm1VkVUpyJ_ur28zblHPwGbGb7oMSaEIVyE5ATIQ=.62df58ce-a6e9-4b01-a5dd-835b66b0c3bb@github.com>

> The `MethodHandles::zero` and `MethodHandles::empty` APIs are updated to return a primitive object with the default value if the given type is a primitive value type. These methods return null if the given type is a value class and a primitive reference type.
> 
> The MethodHandles::constant method will throw `NullPointerException` if the given type is a primitive value type and the given value is null.   That's the same behavior as the `int` type but such case is missing in the spec.

Mandy Chung has updated the pull request incrementally with one additional commit since the last revision:

  fix typo

-------------

Changes:
  - all: https://git.openjdk.java.net/valhalla/pull/644/files
  - new: https://git.openjdk.java.net/valhalla/pull/644/files/e4d0bbe7..5f353636

Webrevs:
 - full: https://webrevs.openjdk.java.net/?repo=valhalla&pr=644&range=01
 - incr: https://webrevs.openjdk.java.net/?repo=valhalla&pr=644&range=00-01

  Stats: 1 line in 1 file changed: 0 ins; 0 del; 1 mod
  Patch: https://git.openjdk.java.net/valhalla/pull/644.diff
  Fetch: git fetch https://git.openjdk.java.net/valhalla pull/644/head:pull/644

PR: https://git.openjdk.java.net/valhalla/pull/644

From mchung at openjdk.java.net  Thu Feb 10 18:13:26 2022
From: mchung at openjdk.java.net (Mandy Chung)
Date: Thu, 10 Feb 2022 18:13:26 GMT
Subject: [lworld] RFR: 8281073: [lworld] Update MethodHandles::zero, empty,
 constant APIs for primitive classes [v2]
In-Reply-To: <RxFmJ7iy_C-RNqj5YCYe5xBLAUs42t_A4WxoFe7TvpQ=.7a7ca52b-5b2f-4395-8065-7d8c0f19055b@github.com>
References: <XgZFb4_E4GhjfuXo8DPw6bmynBOJMDmGIqwBKMR2vYQ=.b828f472-c86f-427e-8d42-1fc198c87899@github.com>
 <RxFmJ7iy_C-RNqj5YCYe5xBLAUs42t_A4WxoFe7TvpQ=.7a7ca52b-5b2f-4395-8065-7d8c0f19055b@github.com>
Message-ID: <W0QhTnqZHTVXy9pZqN16DAL58ObXaxOY_fAvpQ-hJbk=.21cc880e-5708-4ba0-bd83-155bc7a6d2e6@github.com>

On Thu, 10 Feb 2022 15:05:47 GMT, Alex Menkov <amenkov at openjdk.org> wrote:

>> Mandy Chung has updated the pull request incrementally with one additional commit since the last revision:
>> 
>>   fix typo
>
> src/java.base/share/classes/java/lang/invoke/MethodHandles.java line 5116:
> 
>> 5114:      * and returns a suitable default depending on the return type.
>> 5115:      * If the requested type is a primitive type or {@code void}, it returns
>> 5116:      * a zero primitive value or {@ocde void}.
> 
> {@ocde void} -> {@code void}

Thanks for catching the typo.   Fixed.

-------------

PR: https://git.openjdk.java.net/valhalla/pull/644

From hseigel at openjdk.java.net  Thu Feb 10 18:16:35 2022
From: hseigel at openjdk.java.net (Harold Seigel)
Date: Thu, 10 Feb 2022 18:16:35 GMT
Subject: [lworld] RFR: 8281471: [lworld] check_code.c should not accept
 Q-signatures as valid
Message-ID: <cWPCuql-6b_UjLm2wKYW9-ZFxnBybPcqzOkwgDWzupc=.e3bd3b46-6d4d-4654-8392-741f57638d09@github.com>

Please review this small change to not allow Q signatures in old class files that are verified using the old verifier.  The change removes checks for Q signatures from the old verifier.

The fix was tested with Mach5 tiers 1-2 on Linux, Mac OS, and Windows, and Mach5 tiers 3-5 on Linux x64.

Thanks, Harold

-------------

Commit messages:
 - 8281471: [lworld] check_code.c should not accept Q-signatures as valid

Changes: https://git.openjdk.java.net/valhalla/pull/645/files
 Webrev: https://webrevs.openjdk.java.net/?repo=valhalla&pr=645&range=00
  Issue: https://bugs.openjdk.java.net/browse/JDK-8281471
  Stats: 146 lines in 4 files changed: 131 ins; 3 del; 12 mod
  Patch: https://git.openjdk.java.net/valhalla/pull/645.diff
  Fetch: git fetch https://git.openjdk.java.net/valhalla pull/645/head:pull/645

PR: https://git.openjdk.java.net/valhalla/pull/645

From fparain at openjdk.java.net  Thu Feb 10 18:29:38 2022
From: fparain at openjdk.java.net (Frederic Parain)
Date: Thu, 10 Feb 2022 18:29:38 GMT
Subject: [lworld] RFR: 8281471: [lworld] check_code.c should not accept
 Q-signatures as valid
In-Reply-To: <cWPCuql-6b_UjLm2wKYW9-ZFxnBybPcqzOkwgDWzupc=.e3bd3b46-6d4d-4654-8392-741f57638d09@github.com>
References: <cWPCuql-6b_UjLm2wKYW9-ZFxnBybPcqzOkwgDWzupc=.e3bd3b46-6d4d-4654-8392-741f57638d09@github.com>
Message-ID: <lk4C2G5WQOkFOCmsbKDUCS5SAbxIkK09Lov6H4P7sPU=.25a56a23-e691-4b88-85f4-2cece2764323@github.com>

On Thu, 10 Feb 2022 18:10:56 GMT, Harold Seigel <hseigel at openjdk.org> wrote:

> Please review this small change to not allow Q signatures in old class files that are verified using the old verifier.  The change removes checks for Q signatures from the old verifier.
> 
> The fix was tested with Mach5 tiers 1-2 on Linux, Mac OS, and Windows, and Mach5 tiers 3-5 on Linux x64.
> 
> Thanks, Harold

Looks good to me.

Fred

-------------

Marked as reviewed by fparain (Committer).

PR: https://git.openjdk.java.net/valhalla/pull/645

From rriggs at openjdk.java.net  Thu Feb 10 20:52:30 2022
From: rriggs at openjdk.java.net (Roger Riggs)
Date: Thu, 10 Feb 2022 20:52:30 GMT
Subject: [lworld] RFR: 8281073: [lworld] Update MethodHandles::zero, empty,
 constant APIs for primitive classes [v2]
In-Reply-To: <ggKUm1VkVUpyJ_ur28zblHPwGbGb7oMSaEIVyE5ATIQ=.62df58ce-a6e9-4b01-a5dd-835b66b0c3bb@github.com>
References: <XgZFb4_E4GhjfuXo8DPw6bmynBOJMDmGIqwBKMR2vYQ=.b828f472-c86f-427e-8d42-1fc198c87899@github.com>
 <ggKUm1VkVUpyJ_ur28zblHPwGbGb7oMSaEIVyE5ATIQ=.62df58ce-a6e9-4b01-a5dd-835b66b0c3bb@github.com>
Message-ID: <f9Tsj4rdLU_dl9rZ6MmsIwWlF4c1eaM5_7MCcAb9hYM=.2c0878a7-09ce-4878-a236-b536ecdb49f7@github.com>

On Thu, 10 Feb 2022 18:13:25 GMT, Mandy Chung <mchung at openjdk.org> wrote:

>> The `MethodHandles::zero` and `MethodHandles::empty` APIs are updated to return a primitive object with the default value if the given type is a primitive value type. These methods return null if the given type is a value class and a primitive reference type.
>> 
>> The MethodHandles::constant method will throw `NullPointerException` if the given type is a primitive value type and the given value is null.   That's the same behavior as the `int` type but such case is missing in the spec.
>
> Mandy Chung has updated the pull request incrementally with one additional commit since the last revision:
> 
>   fix typo

Marked as reviewed by rriggs (Committer).

-------------

PR: https://git.openjdk.java.net/valhalla/pull/644

From dlsmith at openjdk.java.net  Thu Feb 10 22:01:10 2022
From: dlsmith at openjdk.java.net (Dan Smith)
Date: Thu, 10 Feb 2022 22:01:10 GMT
Subject: [lworld] RFR: 8281618: Add asmtools sources as a test library
Message-ID: <0SUC8DtXWAE2PYQ7SJbh_AIeYeSP3qQP_9rMCT5YrFQ=.6376e86a-58ef-4eec-9817-9705766ba42a@github.com>

This is mostly a direct copy of the asmtools repo. The one interesting file is JtregDriver, which avoids various complications involved in directly invoking the asmtools 'Main' entry point.

-------------

Commit messages:
 - Fix trailing whitespace
 - Added usage comment
 - Generalize JtregDriver for all asmtools commands
 - Add JtregDriver to provide an entry point that doesn't call System.exit
 - Add asmtools source to test/lib

Changes: https://git.openjdk.java.net/valhalla/pull/646/files
 Webrev: https://webrevs.openjdk.java.net/?repo=valhalla&pr=646&range=00
  Issue: https://bugs.openjdk.java.net/browse/JDK-8281618
  Stats: 27569 lines in 101 files changed: 27569 ins; 0 del; 0 mod
  Patch: https://git.openjdk.java.net/valhalla/pull/646.diff
  Fetch: git fetch https://git.openjdk.java.net/valhalla pull/646/head:pull/646

PR: https://git.openjdk.java.net/valhalla/pull/646

From mchung at openjdk.java.net  Thu Feb 10 22:49:46 2022
From: mchung at openjdk.java.net (Mandy Chung)
Date: Thu, 10 Feb 2022 22:49:46 GMT
Subject: [lworld] Integrated: 8281073: [lworld] Update MethodHandles::zero,
 empty, constant APIs for primitive classes
In-Reply-To: <XgZFb4_E4GhjfuXo8DPw6bmynBOJMDmGIqwBKMR2vYQ=.b828f472-c86f-427e-8d42-1fc198c87899@github.com>
References: <XgZFb4_E4GhjfuXo8DPw6bmynBOJMDmGIqwBKMR2vYQ=.b828f472-c86f-427e-8d42-1fc198c87899@github.com>
Message-ID: <4f7iMFFxha0i0CYVVFXAX04D9H70dV2HowwagMbOVG0=.73c3ed81-88da-4235-8523-2a542a826643@github.com>

On Wed, 9 Feb 2022 20:24:07 GMT, Mandy Chung <mchung at openjdk.org> wrote:

> The `MethodHandles::zero` and `MethodHandles::empty` APIs are updated to return a primitive object with the default value if the given type is a primitive value type. These methods return null if the given type is a value class and a primitive reference type.
> 
> The MethodHandles::constant method will throw `NullPointerException` if the given type is a primitive value type and the given value is null.   That's the same behavior as the `int` type but such case is missing in the spec.

This pull request has now been integrated.

Changeset: 611fca0a
Author:    Mandy Chung <mchung at openjdk.org>
URL:       https://git.openjdk.java.net/valhalla/commit/611fca0a13051036144414e0d07509f45b5e70ca
Stats:     132 lines in 2 files changed: 126 ins; 0 del; 6 mod

8281073: [lworld] Update MethodHandles::zero, empty, constant APIs for primitive classes

Reviewed-by: rriggs

-------------

PR: https://git.openjdk.java.net/valhalla/pull/644

From dlsmith at openjdk.java.net  Thu Feb 10 23:56:23 2022
From: dlsmith at openjdk.java.net (Dan Smith)
Date: Thu, 10 Feb 2022 23:56:23 GMT
Subject: [lworld] RFR: 8281618: Add asmtools sources as a test library [v2]
In-Reply-To: <0SUC8DtXWAE2PYQ7SJbh_AIeYeSP3qQP_9rMCT5YrFQ=.6376e86a-58ef-4eec-9817-9705766ba42a@github.com>
References: <0SUC8DtXWAE2PYQ7SJbh_AIeYeSP3qQP_9rMCT5YrFQ=.6376e86a-58ef-4eec-9817-9705766ba42a@github.com>
Message-ID: <SCM1e-zJ0dqMwi9OiCzsWvPAvjgyQnTWn3tCnOMv940=.3d3d646b-d0d4-413e-af69-9c7e966d6bbd@github.com>

> This is mostly a direct copy of the asmtools repo. The one interesting file is JtregDriver, which avoids various complications involved in directly invoking the asmtools 'Main' entry point.

Dan Smith has updated the pull request with a new target base due to a merge or a rebase. The incremental webrev excludes the unrelated changes brought in by the merge/rebase. The pull request contains six additional commits since the last revision:

 - Merge branch 'lworld' into asmtools
 - Fix trailing whitespace
 - Added usage comment
 - Generalize JtregDriver for all asmtools commands
 - Add JtregDriver to provide an entry point that doesn't call System.exit
 - Add asmtools source to test/lib

-------------

Changes:
  - all: https://git.openjdk.java.net/valhalla/pull/646/files
  - new: https://git.openjdk.java.net/valhalla/pull/646/files/ccbec02f..ed43e819

Webrevs:
 - full: https://webrevs.openjdk.java.net/?repo=valhalla&pr=646&range=01
 - incr: https://webrevs.openjdk.java.net/?repo=valhalla&pr=646&range=00-01

  Stats: 106967 lines in 2913 files changed: 70406 ins; 20833 del; 15728 mod
  Patch: https://git.openjdk.java.net/valhalla/pull/646.diff
  Fetch: git fetch https://git.openjdk.java.net/valhalla pull/646/head:pull/646

PR: https://git.openjdk.java.net/valhalla/pull/646

From hseigel at openjdk.java.net  Fri Feb 11 13:26:49 2022
From: hseigel at openjdk.java.net (Harold Seigel)
Date: Fri, 11 Feb 2022 13:26:49 GMT
Subject: [lworld] RFR: 8281471: [lworld] check_code.c should not accept
 Q-signatures as valid
In-Reply-To: <cWPCuql-6b_UjLm2wKYW9-ZFxnBybPcqzOkwgDWzupc=.e3bd3b46-6d4d-4654-8392-741f57638d09@github.com>
References: <cWPCuql-6b_UjLm2wKYW9-ZFxnBybPcqzOkwgDWzupc=.e3bd3b46-6d4d-4654-8392-741f57638d09@github.com>
Message-ID: <4K6aefjUKeuj4hxamZ8vCl_v1yGcpELXkxiSw4559tI=.d4447d3a-089d-4e2c-8b96-07ec83603395@github.com>

On Thu, 10 Feb 2022 18:10:56 GMT, Harold Seigel <hseigel at openjdk.org> wrote:

> Please review this small change to not allow Q signatures in old class files that are verified using the old verifier.  The change removes checks for Q signatures from the old verifier.
> 
> The fix was tested with Mach5 tiers 1-2 on Linux, Mac OS, and Windows, and Mach5 tiers 3-5 on Linux x64.
> 
> Thanks, Harold

Thanks Fred for the review!

-------------

PR: https://git.openjdk.java.net/valhalla/pull/645

From hseigel at openjdk.java.net  Fri Feb 11 13:26:50 2022
From: hseigel at openjdk.java.net (Harold Seigel)
Date: Fri, 11 Feb 2022 13:26:50 GMT
Subject: [lworld] Integrated: 8281471: [lworld] check_code.c should not accept
 Q-signatures as valid
In-Reply-To: <cWPCuql-6b_UjLm2wKYW9-ZFxnBybPcqzOkwgDWzupc=.e3bd3b46-6d4d-4654-8392-741f57638d09@github.com>
References: <cWPCuql-6b_UjLm2wKYW9-ZFxnBybPcqzOkwgDWzupc=.e3bd3b46-6d4d-4654-8392-741f57638d09@github.com>
Message-ID: <EEDif60PWU4fdfXIvUp8vGGmevXmw87OExkDcr74u7k=.6acc45b2-77b7-4328-b318-63eadf8b05b6@github.com>

On Thu, 10 Feb 2022 18:10:56 GMT, Harold Seigel <hseigel at openjdk.org> wrote:

> Please review this small change to not allow Q signatures in old class files that are verified using the old verifier.  The change removes checks for Q signatures from the old verifier.
> 
> The fix was tested with Mach5 tiers 1-2 on Linux, Mac OS, and Windows, and Mach5 tiers 3-5 on Linux x64.
> 
> Thanks, Harold

This pull request has now been integrated.

Changeset: c99c872f
Author:    Harold Seigel <hseigel at openjdk.org>
URL:       https://git.openjdk.java.net/valhalla/commit/c99c872f9fcdcb5874b0ef046b1e645a187dba40
Stats:     146 lines in 4 files changed: 131 ins; 3 del; 12 mod

8281471: [lworld] check_code.c should not accept Q-signatures as valid

Reviewed-by: fparain

-------------

PR: https://git.openjdk.java.net/valhalla/pull/645

From dsimms at openjdk.java.net  Mon Feb 14 19:17:40 2022
From: dsimms at openjdk.java.net (David Simms)
Date: Mon, 14 Feb 2022 19:17:40 GMT
Subject: [lworld] RFR: 8281618: Add asmtools sources as a test library [v2]
In-Reply-To: <SCM1e-zJ0dqMwi9OiCzsWvPAvjgyQnTWn3tCnOMv940=.3d3d646b-d0d4-413e-af69-9c7e966d6bbd@github.com>
References: <0SUC8DtXWAE2PYQ7SJbh_AIeYeSP3qQP_9rMCT5YrFQ=.6376e86a-58ef-4eec-9817-9705766ba42a@github.com>
 <SCM1e-zJ0dqMwi9OiCzsWvPAvjgyQnTWn3tCnOMv940=.3d3d646b-d0d4-413e-af69-9c7e966d6bbd@github.com>
Message-ID: <Fmw2AeWZfyCozTx8pQW63ecP860dnAfhgyFatvK3KZQ=.d7025360-762e-441b-875d-fcc31d334b67@github.com>

On Thu, 10 Feb 2022 23:56:23 GMT, Dan Smith <dlsmith at openjdk.org> wrote:

>> This is mostly a direct copy of the asmtools repo. The one interesting file is JtregDriver, which avoids various complications involved in directly invoking the asmtools 'Main' entry point.
>
> Dan Smith has updated the pull request with a new target base due to a merge or a rebase. The incremental webrev excludes the unrelated changes brought in by the merge/rebase. The pull request contains six additional commits since the last revision:
> 
>  - Merge branch 'lworld' into asmtools
>  - Fix trailing whitespace
>  - Added usage comment
>  - Generalize JtregDriver for all asmtools commands
>  - Add JtregDriver to provide an entry point that doesn't call System.exit
>  - Add asmtools source to test/lib

There is not a good way of adding Valhalla features still in prototyping without having a Valhalla specific copy...looks good

-------------

Marked as reviewed by dsimms (Committer).

PR: https://git.openjdk.java.net/valhalla/pull/646

From dlsmith at openjdk.java.net  Mon Feb 14 20:20:39 2022
From: dlsmith at openjdk.java.net (Dan Smith)
Date: Mon, 14 Feb 2022 20:20:39 GMT
Subject: [lworld] Integrated: 8281618: Add asmtools sources as a test library
In-Reply-To: <0SUC8DtXWAE2PYQ7SJbh_AIeYeSP3qQP_9rMCT5YrFQ=.6376e86a-58ef-4eec-9817-9705766ba42a@github.com>
References: <0SUC8DtXWAE2PYQ7SJbh_AIeYeSP3qQP_9rMCT5YrFQ=.6376e86a-58ef-4eec-9817-9705766ba42a@github.com>
Message-ID: <z9yVHvwWHwoPNRltcNi5wF0dzefjlI-3mux3NUmLtUo=.89215132-7cd7-4e10-9cde-a11df0f89ae9@github.com>

On Thu, 10 Feb 2022 21:43:59 GMT, Dan Smith <dlsmith at openjdk.org> wrote:

> Adding the asmtools sources to test/lib to allow for valhalla feature support before the features become standard. (See https://github.com/openjdk/asmtools).
> 
> Example usage: given a source file Foo.jasm in a test directory, the test would typically say:
> 
> `* @build Foo`
> 
> or maybe
> 
> `* @compile Foo.jasm`
> 
> These lines would be replaced with:
> 
> `* @library /test/lib`
> `* @build org.openjdk.asmtools.* org.openjdk.asmtools.jasm.*`
> `* @run driver org.openjdk.asmtools.JtregDriver jasm Foo.jasm`
> 
> I added the JtregDriver class, which takes care of things like making sure the source and class file paths are in the right place (and System.exit doesn't crash jtreg!). Everything else is a straight 'cp -R' from the asmtools repo; I'll follow up with changes to support the Valhalla class file features.

This pull request has now been integrated.

Changeset: dc6036f6
Author:    Dan Smith <dlsmith at openjdk.org>
URL:       https://git.openjdk.java.net/valhalla/commit/dc6036f627813e5605d93be4c89bc8dfa2006bc5
Stats:     27569 lines in 101 files changed: 27569 ins; 0 del; 0 mod

8281618: Add asmtools sources as a test library

Duplicate copy of asmtools that will support Valhalla class file features. Eventually those features will end up in the official asmtools & jtreg, and we can garbage-collect this fork.

Reviewed-by: dsimms

-------------

PR: https://git.openjdk.java.net/valhalla/pull/646

From dlsmith at openjdk.java.net  Mon Feb 14 20:55:07 2022
From: dlsmith at openjdk.java.net (Dan Smith)
Date: Mon, 14 Feb 2022 20:55:07 GMT
Subject: [lworld] RFR: 8281763: Support Valhalla class file features in
 asmtools
Message-ID: <c-GMBATowBmhOhOwMu2uD9dlUTM13uNC9Lojgp2995c=.947f6728-b595-4fbd-8ff7-17576b5d681e@github.com>

Modifying the Valhalla repo's fork of asmtools to implement Valhalla class file features:

- ACC_VALUE, ACC_PERMITS_VALUE, and ACC_PRIMITIVE 
- Preload attribute 
- aconst_init & withfield instructions 
- Unnamed factory methods 
- Q descriptors

Mainly relevant to the `jasm` tool, but there are also some interactions with `jdis` and `jdec`. No impact on `jcod`.

-------------

Commit messages:
 - Merge branch 'asmtools' into asmtools-valhalla
 - Merge branch 'lworld' into asmtools
 - Modified asmtools files
 - AsmTools updates for Valhalla features
 - Merge branch 'lworld' into asmtools
 - Fix trailing whitespace
 - Added usage comment
 - Generalize JtregDriver for all asmtools commands
 - Add JtregDriver to provide an entry point that doesn't call System.exit
 - Add asmtools source to test/lib

Changes: https://git.openjdk.java.net/valhalla/pull/647/files
 Webrev: https://webrevs.openjdk.java.net/?repo=valhalla&pr=647&range=00
  Issue: https://bugs.openjdk.java.net/browse/JDK-8281763
  Stats: 189 lines in 14 files changed: 165 ins; 6 del; 18 mod
  Patch: https://git.openjdk.java.net/valhalla/pull/647.diff
  Fetch: git fetch https://git.openjdk.java.net/valhalla pull/647/head:pull/647

PR: https://git.openjdk.java.net/valhalla/pull/647

From vromero at openjdk.java.net  Tue Feb 15 05:00:53 2022
From: vromero at openjdk.java.net (Vicente Romero)
Date: Tue, 15 Feb 2022 05:00:53 GMT
Subject: Integrated: cleaning up overload resolution code
Message-ID: <j3zH4Af4BL7ylkrTu5KUBUN86y5kJqiWTs0YF5UaaEk=.8aa68d5f-35f6-4b19-8e20-0a60f028f650@github.com>

Overload resolution cleanup

-------------

Commit messages:
 - cleaning up overload resolution code

Changes: https://git.openjdk.java.net/valhalla/pull/648/files
 Webrev: https://webrevs.openjdk.java.net/?repo=valhalla&pr=648&range=00
  Stats: 29 lines in 3 files changed: 3 ins; 13 del; 13 mod
  Patch: https://git.openjdk.java.net/valhalla/pull/648.diff
  Fetch: git fetch https://git.openjdk.java.net/valhalla pull/648/head:pull/648

PR: https://git.openjdk.java.net/valhalla/pull/648

From vromero at openjdk.java.net  Tue Feb 15 05:00:54 2022
From: vromero at openjdk.java.net (Vicente Romero)
Date: Tue, 15 Feb 2022 05:00:54 GMT
Subject: Integrated: cleaning up overload resolution code
In-Reply-To: <j3zH4Af4BL7ylkrTu5KUBUN86y5kJqiWTs0YF5UaaEk=.8aa68d5f-35f6-4b19-8e20-0a60f028f650@github.com>
References: <j3zH4Af4BL7ylkrTu5KUBUN86y5kJqiWTs0YF5UaaEk=.8aa68d5f-35f6-4b19-8e20-0a60f028f650@github.com>
Message-ID: <w3JanmpvB9LHjpuCPqHbSL5dt-RqTNEJ5d6VT0Q40fg=.48ed3671-6b7b-490c-a9b8-fc574db6de7d@github.com>

On Tue, 15 Feb 2022 04:54:02 GMT, Vicente Romero <vromero at openjdk.org> wrote:

> Overload resolution cleanup

This pull request has now been integrated.

Changeset: f171f25b
Author:    Vicente Romero <vromero at openjdk.org>
URL:       https://git.openjdk.java.net/valhalla/commit/f171f25badfb72a444739c79ad54de7ab7b8aa2f
Stats:     29 lines in 3 files changed: 3 ins; 13 del; 13 mod

cleaning up overload resolution code

-------------

PR: https://git.openjdk.java.net/valhalla/pull/648

From dsimms at openjdk.java.net  Tue Feb 15 08:50:55 2022
From: dsimms at openjdk.java.net (David Simms)
Date: Tue, 15 Feb 2022 08:50:55 GMT
Subject: [lworld] Integrated: Tweak header format for correctness
Message-ID: <ArIrjwDUl5ScsuMbAuDpezoQoLOa3vY2c27BOdwMuKo=.6d3cc0e2-6ef8-4a2a-b6cc-af74b393c9a7@github.com>

Formatting

-------------

Commit messages:
 - Tweak header format for correctness

Changes: https://git.openjdk.java.net/valhalla/pull/649/files
 Webrev: https://webrevs.openjdk.java.net/?repo=valhalla&pr=649&range=00
  Stats: 4 lines in 4 files changed: 0 ins; 0 del; 4 mod
  Patch: https://git.openjdk.java.net/valhalla/pull/649.diff
  Fetch: git fetch https://git.openjdk.java.net/valhalla pull/649/head:pull/649

PR: https://git.openjdk.java.net/valhalla/pull/649

From dsimms at openjdk.java.net  Tue Feb 15 08:50:55 2022
From: dsimms at openjdk.java.net (David Simms)
Date: Tue, 15 Feb 2022 08:50:55 GMT
Subject: [lworld] Integrated: Tweak header format for correctness
In-Reply-To: <ArIrjwDUl5ScsuMbAuDpezoQoLOa3vY2c27BOdwMuKo=.6d3cc0e2-6ef8-4a2a-b6cc-af74b393c9a7@github.com>
References: <ArIrjwDUl5ScsuMbAuDpezoQoLOa3vY2c27BOdwMuKo=.6d3cc0e2-6ef8-4a2a-b6cc-af74b393c9a7@github.com>
Message-ID: <W656uPvA18L-BSh84SyifNZ3wUg6JSadLO2vLmUkC18=.445527d5-7b56-490d-b470-53d15c718bf0@github.com>

On Tue, 15 Feb 2022 08:43:10 GMT, David Simms <dsimms at openjdk.org> wrote:

> Formatting

This pull request has now been integrated.

Changeset: 7c0494a4
Author:    David Simms <dsimms at openjdk.org>
URL:       https://git.openjdk.java.net/valhalla/commit/7c0494a44bf9a8ac29dcf52c5a3a29409b22c03e
Stats:     4 lines in 4 files changed: 0 ins; 0 del; 4 mod

Tweak header format for correctness

-------------

PR: https://git.openjdk.java.net/valhalla/pull/649

From mcimadamore at openjdk.java.net  Tue Feb 15 09:34:39 2022
From: mcimadamore at openjdk.java.net (Maurizio Cimadamore)
Date: Tue, 15 Feb 2022 09:34:39 GMT
Subject: Integrated: cleaning up overload resolution code
In-Reply-To: <j3zH4Af4BL7ylkrTu5KUBUN86y5kJqiWTs0YF5UaaEk=.8aa68d5f-35f6-4b19-8e20-0a60f028f650@github.com>
References: <j3zH4Af4BL7ylkrTu5KUBUN86y5kJqiWTs0YF5UaaEk=.8aa68d5f-35f6-4b19-8e20-0a60f028f650@github.com>
Message-ID: <3OC6gIzCDsnz15C4DuM8kJjLzNLBWdHOZBNAxnNJQIA=.df594dd2-aa1c-472d-a107-aec71568f910@github.com>

On Tue, 15 Feb 2022 04:54:02 GMT, Vicente Romero <vromero at openjdk.org> wrote:

> Overload resolution cleanup

src/jdk.compiler/share/classes/com/sun/tools/javac/comp/Resolve.java line 1058:

> 1056:             boolean anyIsPrimitiveClass = found.isPrimitiveClass() || req.isPrimitiveClass();
> 1057:             return strict ?
> 1058:                     (anyIsUndetVar && anyIsPrimitiveClass ? false :

Could this code be simplified if we hacked directly into the subtyping routine? Note that, I believe that would be necessary anyway, because you have to take into account cases like:

`List<Point> <: List<T?>`

and, also:

`List<Point> <: List<? extends T?>`

So, I think fixing at the level of subtyping/type-containment is what we want here.

-------------

PR: https://git.openjdk.java.net/valhalla/pull/648

From vromero at openjdk.java.net  Tue Feb 15 13:52:19 2022
From: vromero at openjdk.java.net (Vicente Romero)
Date: Tue, 15 Feb 2022 13:52:19 GMT
Subject: Integrated: cleaning up overload resolution code
In-Reply-To: <3OC6gIzCDsnz15C4DuM8kJjLzNLBWdHOZBNAxnNJQIA=.df594dd2-aa1c-472d-a107-aec71568f910@github.com>
References: <j3zH4Af4BL7ylkrTu5KUBUN86y5kJqiWTs0YF5UaaEk=.8aa68d5f-35f6-4b19-8e20-0a60f028f650@github.com>
 <3OC6gIzCDsnz15C4DuM8kJjLzNLBWdHOZBNAxnNJQIA=.df594dd2-aa1c-472d-a107-aec71568f910@github.com>
Message-ID: <k_K3PRNMvH3I7YiwnKC6YTX36olSNAQ_0ToYccSD8HY=.0a74046f-0c98-4294-bb07-f18c06bd5f7e@github.com>

On Tue, 15 Feb 2022 09:30:44 GMT, Maurizio Cimadamore <mcimadamore at openjdk.org> wrote:

>> Overload resolution cleanup
>
> src/jdk.compiler/share/classes/com/sun/tools/javac/comp/Resolve.java line 1058:
> 
>> 1056:             boolean anyIsPrimitiveClass = found.isPrimitiveClass() || req.isPrimitiveClass();
>> 1057:             return strict ?
>> 1058:                     (anyIsUndetVar && anyIsPrimitiveClass ? false :
> 
> Could this code be simplified if we hacked directly into the subtyping routine? Note that, I believe that would be necessary anyway, because you have to take into account cases like:
> 
> `List<Point> <: List<T?>`
> 
> and, also:
> 
> `List<Point> <: List<? extends T?>`
> 
> So, I think fixing at the level of subtyping/type-containment is what we want here.

sure, I will try this

-------------

PR: https://git.openjdk.java.net/valhalla/pull/648

From dlsmith at openjdk.java.net  Tue Feb 15 20:58:04 2022
From: dlsmith at openjdk.java.net (Dan Smith)
Date: Tue, 15 Feb 2022 20:58:04 GMT
Subject: [lworld] RFR: 8281880: AsmTools driver uses wrong implementation
 classes
Message-ID: <cpgmGb5KCfGdhrvcnwZJ4WetD3OMT_Hb0CYMOFDWeUk=.54e4c3ba-de52-42b7-918f-e1246fbb5c31@github.com>

In Valhalla's copy of the AsmTools library, jtreg runs the JtregDriver with a class loader that knows about the original AsmTools bundled with jtreg. As a result, these original classes are used instead of Valhalla's modified implementation. 

Fix: run the driver code in the context of a class loader that will prefer the Valhalla implementation.

-------------

Commit messages:
 - 8281880: AsmTools driver uses wrong implementation classes
 - Merge branch 'lworld' into asmtools
 - Merge branch 'lworld' into asmtools
 - Fix trailing whitespace
 - Added usage comment
 - Generalize JtregDriver for all asmtools commands
 - Add JtregDriver to provide an entry point that doesn't call System.exit
 - Add asmtools source to test/lib

Changes: https://git.openjdk.java.net/valhalla/pull/650/files
 Webrev: https://webrevs.openjdk.java.net/?repo=valhalla&pr=650&range=00
  Issue: https://bugs.openjdk.java.net/browse/JDK-8281880
  Stats: 141 lines in 1 file changed: 80 ins; 20 del; 41 mod
  Patch: https://git.openjdk.java.net/valhalla/pull/650.diff
  Fetch: git fetch https://git.openjdk.java.net/valhalla pull/650/head:pull/650

PR: https://git.openjdk.java.net/valhalla/pull/650

From dlsmith at openjdk.java.net  Tue Feb 15 22:57:22 2022
From: dlsmith at openjdk.java.net (Dan Smith)
Date: Tue, 15 Feb 2022 22:57:22 GMT
Subject: [lworld] Integrated: 8281880: AsmTools driver uses wrong
 implementation classes
In-Reply-To: <cpgmGb5KCfGdhrvcnwZJ4WetD3OMT_Hb0CYMOFDWeUk=.54e4c3ba-de52-42b7-918f-e1246fbb5c31@github.com>
References: <cpgmGb5KCfGdhrvcnwZJ4WetD3OMT_Hb0CYMOFDWeUk=.54e4c3ba-de52-42b7-918f-e1246fbb5c31@github.com>
Message-ID: <-rU9LJJv1Iga2i4bCzjShBSZGHsV6290I0-dvm6AIEg=.4466a1f1-97bd-4aca-a666-60b88a4af407@github.com>

On Tue, 15 Feb 2022 20:51:42 GMT, Dan Smith <dlsmith at openjdk.org> wrote:

> In Valhalla's copy of the AsmTools library, jtreg runs the JtregDriver with a class loader that knows about the original AsmTools bundled with jtreg. As a result, these original classes are used instead of Valhalla's modified implementation. 
> 
> Fix: run the driver code in the context of a class loader that will prefer the Valhalla implementation.

This pull request has now been integrated.

Changeset: 2c8ecdb0
Author:    Dan Smith <dlsmith at openjdk.org>
URL:       https://git.openjdk.java.net/valhalla/commit/2c8ecdb0cd4a3d1a0d699a5fd2cf769692c73bba
Stats:     141 lines in 1 file changed: 80 ins; 20 del; 41 mod

8281880: AsmTools driver uses wrong implementation classes

-------------

PR: https://git.openjdk.java.net/valhalla/pull/650

From dlsmith at openjdk.java.net  Tue Feb 15 23:22:43 2022
From: dlsmith at openjdk.java.net (Dan Smith)
Date: Tue, 15 Feb 2022 23:22:43 GMT
Subject: [lworld] RFR: 8281763: Support Valhalla class file features in
 asmtools [v2]
In-Reply-To: <c-GMBATowBmhOhOwMu2uD9dlUTM13uNC9Lojgp2995c=.947f6728-b595-4fbd-8ff7-17576b5d681e@github.com>
References: <c-GMBATowBmhOhOwMu2uD9dlUTM13uNC9Lojgp2995c=.947f6728-b595-4fbd-8ff7-17576b5d681e@github.com>
Message-ID: <cB2pJn28_AWMyh-tRfSPJU6sTBgRyOwiuwqcLGODMw8=.b7c1f9ee-e295-4b6a-9e6e-829fb569dc53@github.com>

> Modifying the Valhalla repo's fork of asmtools to implement Valhalla class file features:
> 
> - ACC_VALUE, ACC_PERMITS_VALUE, and ACC_PRIMITIVE 
> - Preload attribute 
> - aconst_init & withfield instructions 
> - Unnamed factory methods 
> - Q descriptors
> 
> Mainly relevant to the `jasm` tool, but there are also some interactions with `jdis` and `jdec`. No impact on `jcod`.

Dan Smith has updated the pull request with a new target base due to a merge or a rebase. The pull request now contains 13 commits:

 - Merge branch 'asmtools' into asmtools-valhalla
 - Merge branch 'lworld' into asmtools
 - 8281880: AsmTools driver uses wrong implementation classes
 - Merge branch 'asmtools' into asmtools-valhalla
 - Merge branch 'lworld' into asmtools
 - Modified asmtools files
 - AsmTools updates for Valhalla features
 - Merge branch 'lworld' into asmtools
 - Fix trailing whitespace
 - Added usage comment
 - ... and 3 more: https://git.openjdk.java.net/valhalla/compare/2c8ecdb0...634c132f

-------------

Changes: https://git.openjdk.java.net/valhalla/pull/647/files
 Webrev: https://webrevs.openjdk.java.net/?repo=valhalla&pr=647&range=01
  Stats: 189 lines in 14 files changed: 165 ins; 6 del; 18 mod
  Patch: https://git.openjdk.java.net/valhalla/pull/647.diff
  Fetch: git fetch https://git.openjdk.java.net/valhalla pull/647/head:pull/647

PR: https://git.openjdk.java.net/valhalla/pull/647

From dlsmith at openjdk.java.net  Tue Feb 15 23:22:44 2022
From: dlsmith at openjdk.java.net (Dan Smith)
Date: Tue, 15 Feb 2022 23:22:44 GMT
Subject: [lworld] RFR: 8281763: Support Valhalla class file features in
 asmtools
In-Reply-To: <c-GMBATowBmhOhOwMu2uD9dlUTM13uNC9Lojgp2995c=.947f6728-b595-4fbd-8ff7-17576b5d681e@github.com>
References: <c-GMBATowBmhOhOwMu2uD9dlUTM13uNC9Lojgp2995c=.947f6728-b595-4fbd-8ff7-17576b5d681e@github.com>
Message-ID: <oF6cmcW_agRM4S6dRh3-UGw-BEx3jeciKZl1_xNlOZw=.21f606c9-9b4c-4456-8910-01cc5e921fbb@github.com>

On Mon, 14 Feb 2022 20:48:06 GMT, Dan Smith <dlsmith at openjdk.org> wrote:

> Modifying the Valhalla repo's fork of asmtools to implement Valhalla class file features:
> 
> - ACC_VALUE, ACC_PERMITS_VALUE, and ACC_PRIMITIVE 
> - Preload attribute 
> - aconst_init & withfield instructions 
> - Unnamed factory methods 
> - Q descriptors
> 
> Mainly relevant to the `jasm` tool, but there are also some interactions with `jdis` and `jdec`. No impact on `jcod`.

I don't think we typically write tests for our testing code, but here's an example that can be used to manually sanity-check the behavior:


/**
 * @test
 * @summary Make sure custom jasm works
 * @library /test/lib
 * @build org.openjdk.asmtools.* org.openjdk.asmtools.jasm.*
 * @run driver org.openjdk.asmtools.JtregDriver jasm -strict JasmTest.jasm
 * @run main JasmTest
 */
 
public final primitive value class JasmTest version 62:0 {

    Preload java/lang/String;
    
    private final Field x:I;
    
    public static Method make:"(I)QJasmTest;" stack 2 {
        aconst_init JasmTest;
        iload_0;
        withfield x:I;
        areturn;
    }
    
    public static Method id:"(QJasmTest;)QJasmTest;" stack 1 {
        aload_0;
        areturn;
    }
    
    public static Method fail:"()V" stack 2 {
        new java/lang/RuntimeException;
        dup;
        invokespecial java/lang/RuntimeException.<init>:"()V";
        athrow;
    }

    public static Method main:"([Ljava/lang/String;)V" stack 1 {
        //invokestatic fail:"()V";
        iconst_2;
        invokestatic make:"(I)QJasmTest;";
        pop;
        return;
    }

}

-------------

PR: https://git.openjdk.java.net/valhalla/pull/647

From lkuskov at openjdk.java.net  Wed Feb 16 02:56:25 2022
From: lkuskov at openjdk.java.net (Leonid Kuskov)
Date: Wed, 16 Feb 2022 02:56:25 GMT
Subject: [lworld] RFR: 8281763: Support Valhalla class file features in
 asmtools [v2]
In-Reply-To: <cB2pJn28_AWMyh-tRfSPJU6sTBgRyOwiuwqcLGODMw8=.b7c1f9ee-e295-4b6a-9e6e-829fb569dc53@github.com>
References: <c-GMBATowBmhOhOwMu2uD9dlUTM13uNC9Lojgp2995c=.947f6728-b595-4fbd-8ff7-17576b5d681e@github.com>
 <cB2pJn28_AWMyh-tRfSPJU6sTBgRyOwiuwqcLGODMw8=.b7c1f9ee-e295-4b6a-9e6e-829fb569dc53@github.com>
Message-ID: <ymLYIZ84b1lh7B5wveggsfkvjWJz-WcYlY_AwYYIH-Q=.30a349fd-4ec6-464e-adc1-b3c848d1a830@github.com>

On Tue, 15 Feb 2022 23:22:43 GMT, Dan Smith <dlsmith at openjdk.org> wrote:

>> Modifying the Valhalla repo's fork of asmtools to implement Valhalla class file features:
>> 
>> - ACC_VALUE, ACC_PERMITS_VALUE, and ACC_PRIMITIVE 
>> - Preload attribute 
>> - aconst_init & withfield instructions 
>> - Unnamed factory methods 
>> - Q descriptors
>> 
>> Mainly relevant to the `jasm` tool, but there are also some interactions with `jdis` and `jdec`. No impact on `jcod`.
>
> Dan Smith has updated the pull request with a new target base due to a merge or a rebase. The pull request now contains 13 commits:
> 
>  - Merge branch 'asmtools' into asmtools-valhalla
>  - Merge branch 'lworld' into asmtools
>  - 8281880: AsmTools driver uses wrong implementation classes
>  - Merge branch 'asmtools' into asmtools-valhalla
>  - Merge branch 'lworld' into asmtools
>  - Modified asmtools files
>  - AsmTools updates for Valhalla features
>  - Merge branch 'lworld' into asmtools
>  - Fix trailing whitespace
>  - Added usage comment
>  - ... and 3 more: https://git.openjdk.java.net/valhalla/compare/2c8ecdb0...634c132f

Looks good to me.

-------------

PR: https://git.openjdk.java.net/valhalla/pull/647

From vromero at openjdk.java.net  Wed Feb 16 03:51:46 2022
From: vromero at openjdk.java.net (Vicente Romero)
Date: Wed, 16 Feb 2022 03:51:46 GMT
Subject: RFR: fail during inference if primitive classes are bounds during the
 first overload resolution phase
Message-ID: <dpuf8fCs5my-BcSK5TfYRh5yinQ1RDIW_6a8y80ZLE8=.3c75bc7e-1167-4597-8e14-06c09db1a881@github.com>

Fail during inference if any bound is a primitive class during the first phase

-------------

Commit messages:
 - removing debugging code
 - fail during inference if primitive classes are bounds during the first overload resolution phase

Changes: https://git.openjdk.java.net/valhalla/pull/651/files
 Webrev: https://webrevs.openjdk.java.net/?repo=valhalla&pr=651&range=00
  Stats: 15 lines in 2 files changed: 11 ins; 3 del; 1 mod
  Patch: https://git.openjdk.java.net/valhalla/pull/651.diff
  Fetch: git fetch https://git.openjdk.java.net/valhalla pull/651/head:pull/651

PR: https://git.openjdk.java.net/valhalla/pull/651

From mcimadamore at openjdk.java.net  Wed Feb 16 09:34:39 2022
From: mcimadamore at openjdk.java.net (Maurizio Cimadamore)
Date: Wed, 16 Feb 2022 09:34:39 GMT
Subject: RFR: fail during inference if primitive classes are bounds during
 the first overload resolution phase
In-Reply-To: <dpuf8fCs5my-BcSK5TfYRh5yinQ1RDIW_6a8y80ZLE8=.3c75bc7e-1167-4597-8e14-06c09db1a881@github.com>
References: <dpuf8fCs5my-BcSK5TfYRh5yinQ1RDIW_6a8y80ZLE8=.3c75bc7e-1167-4597-8e14-06c09db1a881@github.com>
Message-ID: <O_IMjns4Hy8QZa1CsihCMBYf8EwVFSzrMsdFR6AOm-s=.6936caae-075b-45e6-922f-32ab06c2ba2f@github.com>

On Wed, 16 Feb 2022 01:06:18 GMT, Vicente Romero <vromero at openjdk.org> wrote:

> Fail during inference if any bound is a primitive class during the first phase

Marked as reviewed by mcimadamore (Committer).

src/jdk.compiler/share/classes/com/sun/tools/javac/comp/Infer.java line 216:

> 214:             deferredAttrContext.complete();
> 215: 
> 216:             /* if boxing is not allowed then, no undetVar should have a bound that is a primitive class, unless the bound

Looks good and very minimal - probably a good starting option for a prototype. Other approaches I thought of include setting "poison" bounds on the inference variable, so as to make subtyping/equality/containment fail.  But it's more complex and less localized.

I believe this might just work. Of course some tests are needed :-)

-------------

PR: https://git.openjdk.java.net/valhalla/pull/651

From vromero at openjdk.java.net  Wed Feb 16 14:15:36 2022
From: vromero at openjdk.java.net (Vicente Romero)
Date: Wed, 16 Feb 2022 14:15:36 GMT
Subject: RFR: fail during inference if primitive classes are bounds during
 the first overload resolution phase
In-Reply-To: <O_IMjns4Hy8QZa1CsihCMBYf8EwVFSzrMsdFR6AOm-s=.6936caae-075b-45e6-922f-32ab06c2ba2f@github.com>
References: <dpuf8fCs5my-BcSK5TfYRh5yinQ1RDIW_6a8y80ZLE8=.3c75bc7e-1167-4597-8e14-06c09db1a881@github.com>
 <O_IMjns4Hy8QZa1CsihCMBYf8EwVFSzrMsdFR6AOm-s=.6936caae-075b-45e6-922f-32ab06c2ba2f@github.com>
Message-ID: <hI8v4p0DursKPgBljPapsZvd8cIv7ygktrFvCV1Hds0=.5fef22eb-61ce-4815-8431-f5f5d373776b@github.com>

On Wed, 16 Feb 2022 09:31:16 GMT, Maurizio Cimadamore <mcimadamore at openjdk.org> wrote:

>> Fail during inference if any bound is a primitive class during the first phase
>
> src/jdk.compiler/share/classes/com/sun/tools/javac/comp/Infer.java line 216:
> 
>> 214:             deferredAttrContext.complete();
>> 215: 
>> 216:             /* if boxing is not allowed then, no undetVar should have a bound that is a primitive class, unless the bound
> 
> Looks good and very minimal - probably a good starting option for a prototype. Other approaches I thought of include setting "poison" bounds on the inference variable, so as to make subtyping/equality/containment fail.  But it's more complex and less localized.
> 
> I believe this might just work. Of course some tests are needed :-)

yep, I will add more tests, thanks for the review

-------------

PR: https://git.openjdk.java.net/valhalla/pull/651

From vromero at openjdk.java.net  Wed Feb 16 14:15:39 2022
From: vromero at openjdk.java.net (Vicente Romero)
Date: Wed, 16 Feb 2022 14:15:39 GMT
Subject: Integrated: fail during inference if primitive classes are bounds
 during the first overload resolution phase
In-Reply-To: <dpuf8fCs5my-BcSK5TfYRh5yinQ1RDIW_6a8y80ZLE8=.3c75bc7e-1167-4597-8e14-06c09db1a881@github.com>
References: <dpuf8fCs5my-BcSK5TfYRh5yinQ1RDIW_6a8y80ZLE8=.3c75bc7e-1167-4597-8e14-06c09db1a881@github.com>
Message-ID: <gTMLMlmpRyQmb3VCQEXdCLQrnPXs5R7rcxInlODp3L8=.b3b1eb8b-2648-4208-bdd6-aa51cf4fea0c@github.com>

On Wed, 16 Feb 2022 01:06:18 GMT, Vicente Romero <vromero at openjdk.org> wrote:

> Fail during inference if any bound is a primitive class during the first phase

This pull request has now been integrated.

Changeset: 07f6c61c
Author:    Vicente Romero <vromero at openjdk.org>
URL:       https://git.openjdk.java.net/valhalla/commit/07f6c61c45a6abc8504d85713fe27bed0a4f38b9
Stats:     15 lines in 2 files changed: 11 ins; 3 del; 1 mod

fail during inference if primitive classes are bounds during the first overload resolution phase

Reviewed-by: mcimadamore

-------------

PR: https://git.openjdk.java.net/valhalla/pull/651

From dlsmith at openjdk.java.net  Wed Feb 16 15:19:10 2022
From: dlsmith at openjdk.java.net (Dan Smith)
Date: Wed, 16 Feb 2022 15:19:10 GMT
Subject: git: openjdk/valhalla: lworld: 8281763: Support Valhalla class file
 features in asmtools
Message-ID: <238e4ce0-619f-4027-885d-d3385b8e5761@openjdk.org>

Changeset: 6c863ce4
Author:    Dan Smith <dlsmith at openjdk.org>
Date:      2022-02-16 15:18:44 +0000
URL:       https://git.openjdk.java.net/valhalla/commit/6c863ce493e37c652406f08f1726a5518cae7116

8281763: Support Valhalla class file features in asmtools

Reviewed-by: lkuskov

! test/lib/org/openjdk/asmtools/jasm/ClassData.java
! test/lib/org/openjdk/asmtools/jasm/JasmTokens.java
! test/lib/org/openjdk/asmtools/jasm/Modifiers.java
! test/lib/org/openjdk/asmtools/jasm/OpcodeTables.java
! test/lib/org/openjdk/asmtools/jasm/Parser.java
! test/lib/org/openjdk/asmtools/jasm/ParserInstr.java
+ test/lib/org/openjdk/asmtools/jasm/PreloadAttr.java
! test/lib/org/openjdk/asmtools/jasm/RuntimeConstants.java
! test/lib/org/openjdk/asmtools/jasm/Tables.java
! test/lib/org/openjdk/asmtools/jasm/i18n.properties
! test/lib/org/openjdk/asmtools/jdec/ClassData.java
! test/lib/org/openjdk/asmtools/jdis/ClassData.java
! test/lib/org/openjdk/asmtools/jdis/CodeData.java
+ test/lib/org/openjdk/asmtools/jdis/PreloadData.java


From dlsmith at openjdk.java.net  Wed Feb 16 15:23:45 2022
From: dlsmith at openjdk.java.net (Dan Smith)
Date: Wed, 16 Feb 2022 15:23:45 GMT
Subject: [lworld] Integrated: 8281763: Support Valhalla class file features in
 asmtools
In-Reply-To: <c-GMBATowBmhOhOwMu2uD9dlUTM13uNC9Lojgp2995c=.947f6728-b595-4fbd-8ff7-17576b5d681e@github.com>
References: <c-GMBATowBmhOhOwMu2uD9dlUTM13uNC9Lojgp2995c=.947f6728-b595-4fbd-8ff7-17576b5d681e@github.com>
Message-ID: <mWeE5c7hfuItOsQaa_D8YkUn--nVyjMWbKZk1Jn4G_M=.79bf82d9-3704-4ef1-bc27-ca58a24d0686@github.com>

On Mon, 14 Feb 2022 20:48:06 GMT, Dan Smith <dlsmith at openjdk.org> wrote:

> Modifying the Valhalla repo's fork of asmtools to implement Valhalla class file features:
> 
> - ACC_VALUE, ACC_PERMITS_VALUE, and ACC_PRIMITIVE 
> - Preload attribute 
> - aconst_init & withfield instructions 
> - Unnamed factory methods 
> - Q descriptors
> 
> Mainly relevant to the `jasm` tool, but there are also some interactions with `jdis` and `jdec`. No impact on `jcod`.

This pull request has now been integrated.

Changeset: 6c863ce4
Author:    Dan Smith <dlsmith at openjdk.org>
URL:       https://git.openjdk.java.net/valhalla/commit/6c863ce493e37c652406f08f1726a5518cae7116
Stats:     189 lines in 14 files changed: 165 ins; 6 del; 18 mod

8281763: Support Valhalla class file features in asmtools

Reviewed-by: lkuskov

-------------

PR: https://git.openjdk.java.net/valhalla/pull/647

From vromero at openjdk.java.net  Thu Feb 17 00:47:33 2022
From: vromero at openjdk.java.net (Vicente Romero)
Date: Thu, 17 Feb 2022 00:47:33 GMT
Subject: Integrated: Merge lworld
Message-ID: <VGkXUfWim2cJocFoVtE66rjImPFXGyswIWfHLM44BoE=.430afb69-69f3-42ec-a02c-158a8c226d01@github.com>

Merge lworld into universal-tvars

-------------

Commit messages:
 - Merge branch 'lworld' into universal-tvars_merge_lworld_1
 - fail during inference if primitive classes are bounds during the first overload resolution phase
 - cleaning up overload resolution code
 - Merge lworld
 - some existing regression tests are failing due to universal tvars changes
 - Adding type system regression tests
 - some bug fixing, adding new regression tests
 - Merge lworld
 - Primitive value conversion
 - updating tests
 - ... and 6 more: https://git.openjdk.java.net/valhalla/compare/6c863ce4...3f6df0aa

The merge commit only contains trivial merges, so no merge-specific webrevs have been generated.

Changes: https://git.openjdk.java.net/valhalla/pull/652/files
  Stats: 1182 lines in 29 files changed: 1079 ins; 26 del; 77 mod
  Patch: https://git.openjdk.java.net/valhalla/pull/652.diff
  Fetch: git fetch https://git.openjdk.java.net/valhalla pull/652/head:pull/652

PR: https://git.openjdk.java.net/valhalla/pull/652

From vromero at openjdk.java.net  Thu Feb 17 00:47:34 2022
From: vromero at openjdk.java.net (Vicente Romero)
Date: Thu, 17 Feb 2022 00:47:34 GMT
Subject: Integrated: Merge lworld
In-Reply-To: <VGkXUfWim2cJocFoVtE66rjImPFXGyswIWfHLM44BoE=.430afb69-69f3-42ec-a02c-158a8c226d01@github.com>
References: <VGkXUfWim2cJocFoVtE66rjImPFXGyswIWfHLM44BoE=.430afb69-69f3-42ec-a02c-158a8c226d01@github.com>
Message-ID: <uwFvewOQZCI7yaQG9YfS3qmgUU0O1T8XMtQRqtrrhvY=.a5af4d1f-d20f-47e4-9f03-641a9811a23e@github.com>

On Thu, 17 Feb 2022 00:34:47 GMT, Vicente Romero <vromero at openjdk.org> wrote:

> Merge lworld into universal-tvars

This pull request has now been integrated.

Changeset: 469c04cb
Author:    Vicente Romero <vromero at openjdk.org>
URL:       https://git.openjdk.java.net/valhalla/commit/469c04cbc2e047cc8065fc7953471a27bc731da8
Stats:     28170 lines in 126 files changed: 28069 ins; 25 del; 76 mod

Merge lworld

-------------

PR: https://git.openjdk.java.net/valhalla/pull/652

From dlsmith at openjdk.java.net  Thu Feb 17 01:03:36 2022
From: dlsmith at openjdk.java.net (Dan Smith)
Date: Thu, 17 Feb 2022 01:03:36 GMT
Subject: [lworld] RFR: 8281283: Remove usages of __WithField in runtime tests
Message-ID: <HyjRamI2444_LAuhOYTpn4wHfNCtVLxvoxG0kdAIx-U=.a88a4b33-56e9-4734-950d-7ca21f4fd593@github.com>

Modifying Valhalla runtime tests to no longer use `__WithField` as a macro for the `withfield` instruction. Most uses can be replaced with standard constructor compilation via javac. The remaining uses are rewritten in jasm.

Summary of changes:

- Remove all uses of `-XDallowWithFieldOperator`
- Replace various `make` and `create` factory methods for primitive classes with constructors
- Bug fixes for `JumboInline` that were exposed by this change
- Rewrote `TestFieldNullability` to use jasm
- Rewrote `VWithFieldTest` to use jasm, renamed `WithFieldTest`
- Rewrote `WithFieldAccessorTest` to use jasm, made more comprehensive
- Deleted `WithFieldNoAccessTest`, made redundant by `WithFieldAccessorTest` changes
- Minor tweaks to `TestFieldTypeMismatch` for consistent jasm usage
- Cleanup of some debugging output in `asmtools.JtregDriver`

-------------

Commit messages:
 - Fix whitespace
 - WithFieldNoAccessTest is superceded by revised WithFieldAccessTest
 - withfield tests using jasm
 - Fix up jasm tests
 - Fix whitespace
 - Merge branch 'lworld' into 8281283
 - Convert TestFieldNullability to jasm
 - Convert __Withfield usages in factories to constructors

Changes: https://git.openjdk.java.net/valhalla/pull/653/files
 Webrev: https://webrevs.openjdk.java.net/?repo=valhalla&pr=653&range=00
  Issue: https://bugs.openjdk.java.net/browse/JDK-8281283
  Stats: 2016 lines in 38 files changed: 703 ins; 1105 del; 208 mod
  Patch: https://git.openjdk.java.net/valhalla/pull/653.diff
  Fetch: git fetch https://git.openjdk.java.net/valhalla pull/653/head:pull/653

PR: https://git.openjdk.java.net/valhalla/pull/653

From duke at openjdk.java.net  Thu Feb 17 12:03:36 2022
From: duke at openjdk.java.net (ExE Boss)
Date: Thu, 17 Feb 2022 12:03:36 GMT
Subject: [lworld] RFR: 8281463: [lworld] VALUE / PRIMITIVE modifiers should
 be supported by reflection
In-Reply-To: <H11q_a517sX2Ta6V9p9meFxazzPRhYO7xtKon6DhrVQ=.17fc9f8d-05ad-4b32-ae82-30f4650ee13b@github.com>
References: <H11q_a517sX2Ta6V9p9meFxazzPRhYO7xtKon6DhrVQ=.17fc9f8d-05ad-4b32-ae82-30f4650ee13b@github.com>
Message-ID: <ID2e4QhcqNTjVC8KL0TK9seNiTU1fRhjqiInIxpfZ1o=.e1a1f819-34ac-4719-b003-43dee5e43107@github.com>

On Tue, 8 Feb 2022 21:23:30 GMT, Roger Riggs <rriggs at openjdk.org> wrote:

> Update java.lang.reflection.Modifier to reflect the JVMS specified bits in the modifier.
> 
> The re-use of bits between method and class modifiers poses a problem for this API.
> Previously, all bits were distinct and the modifiers do not include any indication of whether
> the modifier bits are from a class, field, or method. As a result there is ambiguity as two which
> of two meanings a bit takes on.  NATIVE vs VALUE and STRICT vs PRIMITIVE.
> 
> Please review for discussion purposes.

I?personally?think that?extending the?upcoming `AccessFlag`[^1]?enum is?a?better?option.

[^1]: https://github.com/openjdk/jdk/pull/7445

-------------

PR: https://git.openjdk.java.net/valhalla/pull/640

From dlsmith at openjdk.java.net  Thu Feb 17 19:13:41 2022
From: dlsmith at openjdk.java.net (Dan Smith)
Date: Thu, 17 Feb 2022 19:13:41 GMT
Subject: [lworld] RFR: 8281283: Remove usages of __WithField in runtime
 tests
In-Reply-To: <HyjRamI2444_LAuhOYTpn4wHfNCtVLxvoxG0kdAIx-U=.a88a4b33-56e9-4734-950d-7ca21f4fd593@github.com>
References: <HyjRamI2444_LAuhOYTpn4wHfNCtVLxvoxG0kdAIx-U=.a88a4b33-56e9-4734-950d-7ca21f4fd593@github.com>
Message-ID: <sU9NZuE3vLdw8FgqP2rAA3ehciLVt7BVbCUJIv6Cns8=.b1965094-f587-49f6-912f-0d71b1e755ca@github.com>

On Thu, 17 Feb 2022 00:56:50 GMT, Dan Smith <dlsmith at openjdk.org> wrote:

> Modifying Valhalla runtime tests to no longer use `__WithField` as a macro for the `withfield` instruction. Most uses can be replaced with standard constructor compilation via javac. The remaining uses are rewritten in jasm.
> 
> Summary of changes:
> 
> - Remove all uses of `-XDallowWithFieldOperator`
> - Replace various `make` and `create` factory methods for primitive classes with constructors
> - Bug fixes for `JumboInline` that were exposed by this change
> - Rewrote `TestFieldNullability` to use jasm
> - Rewrote `VWithFieldTest` to use jasm, renamed `WithFieldTest`
> - Rewrote `WithFieldAccessorTest` to use jasm, made more comprehensive
> - Deleted `WithFieldNoAccessTest`, made redundant by `WithFieldAccessorTest` changes
> - Minor tweaks to `TestFieldTypeMismatch` for consistent jasm usage
> - Cleanup of some debugging output in `asmtools.JtregDriver`

test/hotspot/jtreg/runtime/valhalla/inlinetypes/JumboInline.java line 79:

> 77:         j = __WithField(j.l1, l1);
> 78:         j = __WithField(j.l2, l0 + l1);
> 79:         j = __WithField(j.l3, l1 + l2);

Bug here (and the following lines): `l2` refers to `this.l2`, not `j.l2`?if `this` is `JumboInline.default`, that value is 0, not `l0+l1`.

test/hotspot/jtreg/runtime/valhalla/inlinetypes/JumboInline.java line 113:

> 111:             return (l0 == j.l0 && l1 == j.l1 && l2 == j.l2 && l3 == j.l3
> 112:                     && l4 == j.l4 && l5 == j.l5 && l6 == j.l6 && l7 == j.l7
> 113:                     && l8 == j.l8 && l9 == j.l9 && l10 == j.l10 && l7 == j.l10

`l7 == j.l10`? Happens to work, because these are always 0, per the above bug.

-------------

PR: https://git.openjdk.java.net/valhalla/pull/653

From dlsmith at openjdk.java.net  Thu Feb 17 19:26:29 2022
From: dlsmith at openjdk.java.net (Dan Smith)
Date: Thu, 17 Feb 2022 19:26:29 GMT
Subject: [lworld] RFR: 8281283: Remove usages of __WithField in runtime
 tests
In-Reply-To: <HyjRamI2444_LAuhOYTpn4wHfNCtVLxvoxG0kdAIx-U=.a88a4b33-56e9-4734-950d-7ca21f4fd593@github.com>
References: <HyjRamI2444_LAuhOYTpn4wHfNCtVLxvoxG0kdAIx-U=.a88a4b33-56e9-4734-950d-7ca21f4fd593@github.com>
Message-ID: <St2lOUgVrFM7kX6ONbAkZER4ks81idHbzsjI2CbMcVk=.e8b35807-e2b8-4180-b37a-440a1073fbd8@github.com>

On Thu, 17 Feb 2022 00:56:50 GMT, Dan Smith <dlsmith at openjdk.org> wrote:

> Modifying Valhalla runtime tests to no longer use `__WithField` as a macro for the `withfield` instruction. Most uses can be replaced with standard constructor compilation via javac. The remaining uses are rewritten in jasm.
> 
> Summary of changes:
> 
> - Remove all uses of `-XDallowWithFieldOperator`
> - Replace various `make` and `create` factory methods for primitive classes with constructors
> - Bug fixes for `JumboInline` that were exposed by this change
> - Rewrote `TestFieldNullability` to use jasm
> - Rewrote `VWithFieldTest` to use jasm, renamed `WithFieldTest`
> - Rewrote `WithFieldAccessorTest` to use jasm, made more comprehensive
> - Deleted `WithFieldNoAccessTest`, made redundant by `WithFieldAccessorTest` changes
> - Minor tweaks to `TestFieldTypeMismatch` for consistent jasm usage
> - Cleanup of some debugging output in `asmtools.JtregDriver`

test/hotspot/jtreg/runtime/valhalla/inlinetypes/TestFieldTypeMismatch.java line 32:

> 30:  * @library /test/lib
> 31:  * @build org.openjdk.asmtools.* org.openjdk.asmtools.jasm.*
> 32:  * @run driver org.openjdk.asmtools.JtregDriver jasm -strict TestFieldTypeMismatchClasses.jasm

Naming convention change: suggest putting jasm classes in a file named `[Test name]Classes.jasm`. No need for any classes declared there to actually use that name, although they should have names that won't clash with other tests in this directory.

test/hotspot/jtreg/runtime/valhalla/inlinetypes/TestValue4.java line 48:

> 46: 
> 47:     public TestValue4() {
> 48:         this((int) System.nanoTime());

Don't need the bit twiddling logic, because it's reproduced by the ByteBuffer reads in the other constructor.

test/hotspot/jtreg/runtime/valhalla/inlinetypes/WithFieldAccessorTest.java line 66:

> 64:         catchAccessError(() -> WithFieldSamePackage.withL(start, 10));
> 65:         catchAccessError(() -> WithFieldSamePackage.withD(start, 11));
> 66:         catchAccessError(() -> WithFieldSamePackage.withI(start, 12));

These calls replace the WithFieldNoAccessTest. (Not checking error messages. Should I add that?)

test/hotspot/jtreg/runtime/valhalla/inlinetypes/WithFieldAccessorTestClasses.jasm line 31:

> 29:     public final Field c:C;
> 30:     protected final Field l:J;
> 31:     final Field d:D;

Added a package-access field to test that, too.

test/hotspot/jtreg/runtime/valhalla/inlinetypes/WithFieldAccessorTestClasses.jasm line 151:

> 149: }
> 150: 
> 151: public final primitive value class runtime/valhalla/inlinetypes/WithFieldNestHost

Additional test methods here, so we can test both the nest "sibling" and the nest "parent".

test/hotspot/jtreg/runtime/valhalla/inlinetypes/WithFieldTest.java line 1:

> 1: /*

Renamed VWithFieldTest --> WithFieldTest

test/hotspot/jtreg/runtime/valhalla/inlinetypes/WithFieldTest.java line 77:

> 75:     static Point[] p_values = new Point[] { new Point(0, 0), new Point(-1, 1), new Point(1, -1)};
> 76: 
> 77:     static void allTypesTest() {

Restructured this logic to avoid managing a bunch of arrays in jasm. Same behavior, the test data just lives outside of the class now.

-------------

PR: https://git.openjdk.java.net/valhalla/pull/653

From duke at openjdk.java.net  Thu Feb 17 19:35:35 2022
From: duke at openjdk.java.net (liach)
Date: Thu, 17 Feb 2022 19:35:35 GMT
Subject: [lworld] RFR: 8281463: [lworld] VALUE / PRIMITIVE modifiers should
 be supported by reflection
In-Reply-To: <H11q_a517sX2Ta6V9p9meFxazzPRhYO7xtKon6DhrVQ=.17fc9f8d-05ad-4b32-ae82-30f4650ee13b@github.com>
References: <H11q_a517sX2Ta6V9p9meFxazzPRhYO7xtKon6DhrVQ=.17fc9f8d-05ad-4b32-ae82-30f4650ee13b@github.com>
Message-ID: <8E6nSQK5AAnFJ4YuTG4aA51xlko_5qgqZZrpzcuD_kw=.468aa2e7-b506-44e4-b21f-318129406106@github.com>

On Tue, 8 Feb 2022 21:23:30 GMT, Roger Riggs <rriggs at openjdk.org> wrote:

> Update java.lang.reflection.Modifier to reflect the JVMS specified bits in the modifier.
> 
> The re-use of bits between method and class modifiers poses a problem for this API.
> Previously, all bits were distinct and the modifiers do not include any indication of whether
> the modifier bits are from a class, field, or method. As a result there is ambiguity as two which
> of two meanings a bit takes on.  NATIVE vs VALUE and STRICT vs PRIMITIVE.
> 
> Please review for discussion purposes.

> I?personally?think that?extending the?upcoming `AccessFlag`?enum is?a?better?option.

I recall that pr was created exactly for valhalla: that in valhalla, class modifier start to use bits used by method/parameter-exclusive modifiers, which made that issue more urgent.

-------------

PR: https://git.openjdk.java.net/valhalla/pull/640

From rriggs at openjdk.java.net  Thu Feb 17 19:56:35 2022
From: rriggs at openjdk.java.net (Roger Riggs)
Date: Thu, 17 Feb 2022 19:56:35 GMT
Subject: [lworld] RFR: 8281463: [lworld] VALUE / PRIMITIVE modifiers should
 be supported by reflection
In-Reply-To: <H11q_a517sX2Ta6V9p9meFxazzPRhYO7xtKon6DhrVQ=.17fc9f8d-05ad-4b32-ae82-30f4650ee13b@github.com>
References: <H11q_a517sX2Ta6V9p9meFxazzPRhYO7xtKon6DhrVQ=.17fc9f8d-05ad-4b32-ae82-30f4650ee13b@github.com>
Message-ID: <CnTwYWLK5Nx2tVTVqRwt4-6AAlckB4rdxqUnwHV9NX8=.0dbe263d-e4bc-48b9-b250-e7e670b8e464@github.com>

On Tue, 8 Feb 2022 21:23:30 GMT, Roger Riggs <rriggs at openjdk.org> wrote:

> Update java.lang.reflection.Modifier to reflect the JVMS specified bits in the modifier.
> 
> The re-use of bits between method and class modifiers poses a problem for this API.
> Previously, all bits were distinct and the modifiers do not include any indication of whether
> the modifier bits are from a class, field, or method. As a result there is ambiguity as two which
> of two meanings a bit takes on.  NATIVE vs VALUE and STRICT vs PRIMITIVE.
> 
> Please review for discussion purposes.

The need in Valhalla isn't urgent and the need prompted the AccessFlag proposal.
I figure to wait for that to settle and be integrated into Valhalla.
Then revise this PR to refer people to AccessFlags and update the javadoc only to avoid confusion.
I'll switch this back to "Draft" for the time being.

-------------

PR: https://git.openjdk.java.net/valhalla/pull/640

From vromero at openjdk.java.net  Fri Feb 18 01:02:42 2022
From: vromero at openjdk.java.net (Vicente Romero)
Date: Fri, 18 Feb 2022 01:02:42 GMT
Subject: RFR: fixing bugs in inference and type system more regression tests
Message-ID: <Nk5CWDiE4Eda7NevLn4bUVMWSzpSDQw-1ZW5HJG_kZw=.65ba3490-e13b-4798-9bcd-35ebf1db3ee2@github.com>

this PR is adapting Types::glb so that it can deal with primitive classes. It also fixes a bug in inference, basically we are now failing during overload resolution if any undetVar has a bound that is a primitive class. But a bug was making inference fail after overload resolution in some cases.

-------------

Commit messages:
 - fixing bugs in inference and type system

Changes: https://git.openjdk.java.net/valhalla/pull/654/files
 Webrev: https://webrevs.openjdk.java.net/?repo=valhalla&pr=654&range=00
  Stats: 36 lines in 3 files changed: 31 ins; 1 del; 4 mod
  Patch: https://git.openjdk.java.net/valhalla/pull/654.diff
  Fetch: git fetch https://git.openjdk.java.net/valhalla pull/654/head:pull/654

PR: https://git.openjdk.java.net/valhalla/pull/654

From mcimadamore at openjdk.java.net  Fri Feb 18 11:09:27 2022
From: mcimadamore at openjdk.java.net (Maurizio Cimadamore)
Date: Fri, 18 Feb 2022 11:09:27 GMT
Subject: RFR: fixing bugs in inference and type system more regression
 tests
In-Reply-To: <Nk5CWDiE4Eda7NevLn4bUVMWSzpSDQw-1ZW5HJG_kZw=.65ba3490-e13b-4798-9bcd-35ebf1db3ee2@github.com>
References: <Nk5CWDiE4Eda7NevLn4bUVMWSzpSDQw-1ZW5HJG_kZw=.65ba3490-e13b-4798-9bcd-35ebf1db3ee2@github.com>
Message-ID: <LyCpXcgDw8arYyRIli3Oe4uZPXDfjcyj5f12eAI4TyA=.51ab2d99-e74c-414c-a8f9-ea9cd26d0b13@github.com>

On Fri, 18 Feb 2022 00:55:43 GMT, Vicente Romero <vromero at openjdk.org> wrote:

> this PR is adapting Types::glb so that it can deal with primitive classes. It also fixes a bug in inference, basically we are now failing during overload resolution if any undetVar has a bound that is a primitive class. But a bug was making inference fail after overload resolution in some cases.

Looks good. I think we should add more tests - e.g. you want to check that, given two primitive types P1 and P2 which both extend a common superinterface I, then:
* lub(P1, P2) == lub(P1, I) == lub(P2, I) == I.
* lub(P1, P1.ref) = P1.ref
* glb(P1, P1.ref) = P1
* glb(P1, I) == P1
* glb(P1.ref, I) == P1.ref
* glb(P2, I) == P2
* glb(P2.ref, I) == P2.ref

-------------

PR: https://git.openjdk.java.net/valhalla/pull/654

From vromero at openjdk.java.net  Fri Feb 18 13:56:14 2022
From: vromero at openjdk.java.net (Vicente Romero)
Date: Fri, 18 Feb 2022 13:56:14 GMT
Subject: RFR: fixing bugs in inference and type system more regression
 tests
In-Reply-To: <LyCpXcgDw8arYyRIli3Oe4uZPXDfjcyj5f12eAI4TyA=.51ab2d99-e74c-414c-a8f9-ea9cd26d0b13@github.com>
References: <Nk5CWDiE4Eda7NevLn4bUVMWSzpSDQw-1ZW5HJG_kZw=.65ba3490-e13b-4798-9bcd-35ebf1db3ee2@github.com>
 <LyCpXcgDw8arYyRIli3Oe4uZPXDfjcyj5f12eAI4TyA=.51ab2d99-e74c-414c-a8f9-ea9cd26d0b13@github.com>
Message-ID: <26wH9i7HxeUscq3UsmtohJ0qSXYdiZZkVIlO3zi4mu0=.d8696547-93b6-47be-bafc-ade5301b5a07@github.com>

On Fri, 18 Feb 2022 11:06:24 GMT, Maurizio Cimadamore <mcimadamore at openjdk.org> wrote:

> Looks good. I think we should add more tests - e.g. you want to check that, given two primitive types P1 and P2 which both extend a common superinterface I, then:
> 
>     * lub(P1, P2) == lub(P1, I) == lub(P2, I) == I.
> 
>     * lub(P1, P1.ref) = P1.ref
> 
>     * glb(P1, P1.ref) = P1
> 
>     * glb(P1, I) == P1
> 
>     * glb(P1.ref, I) == P1.ref
> 
>     * glb(P2, I) == P2
> 
>     * glb(P2.ref, I) == P2.ref



> :

sure, I will add more tests before pushing, thanks

-------------

PR: https://git.openjdk.java.net/valhalla/pull/654

From vromero at openjdk.java.net  Fri Feb 18 17:22:00 2022
From: vromero at openjdk.java.net (Vicente Romero)
Date: Fri, 18 Feb 2022 17:22:00 GMT
Subject: RFR: fixing bugs in inference and type system more regression
 tests [v2]
In-Reply-To: <Nk5CWDiE4Eda7NevLn4bUVMWSzpSDQw-1ZW5HJG_kZw=.65ba3490-e13b-4798-9bcd-35ebf1db3ee2@github.com>
References: <Nk5CWDiE4Eda7NevLn4bUVMWSzpSDQw-1ZW5HJG_kZw=.65ba3490-e13b-4798-9bcd-35ebf1db3ee2@github.com>
Message-ID: <nIESSxuV4z3vX_kNo6dxwLuK8DlfmXMPw6AAQM_tc5M=.76db4c02-19aa-404d-9ff9-f7b80a50f685@github.com>

> this PR is adapting Types::glb so that it can deal with primitive classes. It also fixes a bug in inference, basically we are now failing during overload resolution if any undetVar has a bound that is a primitive class. But a bug was making inference fail after overload resolution in some cases.

Vicente Romero has updated the pull request incrementally with two additional commits since the last revision:

 - Merge remote-tracking branch 'refs/remotes/origin/fixing.bugs.in.inference' into fixing.bugs.in.inference
 - fixing bugs in inference and type system

-------------

Changes:
  - all: https://git.openjdk.java.net/valhalla/pull/654/files
  - new: https://git.openjdk.java.net/valhalla/pull/654/files/860c6726..00ad577c

Webrevs:
 - full: https://webrevs.openjdk.java.net/?repo=valhalla&pr=654&range=01
 - incr: https://webrevs.openjdk.java.net/?repo=valhalla&pr=654&range=00-01

  Stats: 8 lines in 1 file changed: 5 ins; 0 del; 3 mod
  Patch: https://git.openjdk.java.net/valhalla/pull/654.diff
  Fetch: git fetch https://git.openjdk.java.net/valhalla pull/654/head:pull/654

PR: https://git.openjdk.java.net/valhalla/pull/654

From vromero at openjdk.java.net  Fri Feb 18 17:26:16 2022
From: vromero at openjdk.java.net (Vicente Romero)
Date: Fri, 18 Feb 2022 17:26:16 GMT
Subject: Integrated: fixing bugs in inference and type system more regression
 tests
In-Reply-To: <Nk5CWDiE4Eda7NevLn4bUVMWSzpSDQw-1ZW5HJG_kZw=.65ba3490-e13b-4798-9bcd-35ebf1db3ee2@github.com>
References: <Nk5CWDiE4Eda7NevLn4bUVMWSzpSDQw-1ZW5HJG_kZw=.65ba3490-e13b-4798-9bcd-35ebf1db3ee2@github.com>
Message-ID: <ALcdulHbbn4-FIcH_YLedo_7reDMljx-kMSHSCM1xCk=.28230b75-4e7d-4d2c-8178-b4365e89e7d5@github.com>

On Fri, 18 Feb 2022 00:55:43 GMT, Vicente Romero <vromero at openjdk.org> wrote:

> this PR is adapting Types::glb so that it can deal with primitive classes. It also fixes a bug in inference, basically we are now failing during overload resolution if any undetVar has a bound that is a primitive class. But a bug was making inference fail after overload resolution in some cases.

This pull request has now been integrated.

Changeset: aef8323a
Author:    Vicente Romero <vromero at openjdk.org>
URL:       https://git.openjdk.java.net/valhalla/commit/aef8323a3355b1ed3f00a07907054fca8ea8339f
Stats:     44 lines in 4 files changed: 36 ins; 1 del; 7 mod

fixing bugs in inference and type system more regression tests

-------------

PR: https://git.openjdk.java.net/valhalla/pull/654

From rriggs at openjdk.java.net  Fri Feb 18 21:43:23 2022
From: rriggs at openjdk.java.net (Roger Riggs)
Date: Fri, 18 Feb 2022 21:43:23 GMT
Subject: [lworld] RFR: 8281283: Remove usages of __WithField in runtime
 tests
In-Reply-To: <HyjRamI2444_LAuhOYTpn4wHfNCtVLxvoxG0kdAIx-U=.a88a4b33-56e9-4734-950d-7ca21f4fd593@github.com>
References: <HyjRamI2444_LAuhOYTpn4wHfNCtVLxvoxG0kdAIx-U=.a88a4b33-56e9-4734-950d-7ca21f4fd593@github.com>
Message-ID: <Ie09-ZpkdjY3wIWK63yL0F1BVlE9rRydE2YrxBKw64U=.31e87fc6-44a1-48d3-ae22-f30c02d492d2@github.com>

On Thu, 17 Feb 2022 00:56:50 GMT, Dan Smith <dlsmith at openjdk.org> wrote:

> Modifying Valhalla runtime tests to no longer use `__WithField` as a macro for the `withfield` instruction. Most uses can be replaced with standard constructor compilation via javac. The remaining uses are rewritten in jasm.
> 
> Summary of changes:
> 
> - Remove all uses of `-XDallowWithFieldOperator`
> - Replace various `make` and `create` factory methods for primitive classes with constructors
> - Bug fixes for `JumboInline` that were exposed by this change
> - Rewrote `TestFieldNullability` to use jasm
> - Rewrote `VWithFieldTest` to use jasm, renamed `WithFieldTest`
> - Rewrote `WithFieldAccessorTest` to use jasm, made more comprehensive
> - Deleted `WithFieldNoAccessTest`, made redundant by `WithFieldAccessorTest` changes
> - Minor tweaks to `TestFieldTypeMismatch` for consistent jasm usage
> - Cleanup of some debugging output in `asmtools.JtregDriver`

Spot checked.  LGTM

-------------

Marked as reviewed by rriggs (Committer).

PR: https://git.openjdk.java.net/valhalla/pull/653

From vromero at openjdk.java.net  Fri Feb 18 22:58:41 2022
From: vromero at openjdk.java.net (Vicente Romero)
Date: Fri, 18 Feb 2022 22:58:41 GMT
Subject: Integrated: clean up to warnings related code
Message-ID: <NCSRoaxfvUvbaXU3Quo5ywfa9GQpxsi8cG9OrIPU_rw=.62bad07a-657f-4b77-a93b-344c1d2d3207@github.com>

clean up to warnings for universal tvars

-------------

Commit messages:
 - cleaning up warnings related code

Changes: https://git.openjdk.java.net/valhalla/pull/655/files
 Webrev: https://webrevs.openjdk.java.net/?repo=valhalla&pr=655&range=00
  Stats: 71 lines in 5 files changed: 63 ins; 5 del; 3 mod
  Patch: https://git.openjdk.java.net/valhalla/pull/655.diff
  Fetch: git fetch https://git.openjdk.java.net/valhalla pull/655/head:pull/655

PR: https://git.openjdk.java.net/valhalla/pull/655

From vromero at openjdk.java.net  Fri Feb 18 22:58:42 2022
From: vromero at openjdk.java.net (Vicente Romero)
Date: Fri, 18 Feb 2022 22:58:42 GMT
Subject: Integrated: clean up to warnings related code
In-Reply-To: <NCSRoaxfvUvbaXU3Quo5ywfa9GQpxsi8cG9OrIPU_rw=.62bad07a-657f-4b77-a93b-344c1d2d3207@github.com>
References: <NCSRoaxfvUvbaXU3Quo5ywfa9GQpxsi8cG9OrIPU_rw=.62bad07a-657f-4b77-a93b-344c1d2d3207@github.com>
Message-ID: <5pBYjUFLqsJaxgHSClk5ELJQ0VsJZfYzdqrWgOcYWvk=.3292bf71-71d2-4c7d-9b69-7a862466dbe2@github.com>

On Fri, 18 Feb 2022 22:52:41 GMT, Vicente Romero <vromero at openjdk.org> wrote:

> clean up to warnings for universal tvars

This pull request has now been integrated.

Changeset: b0450ed2
Author:    Vicente Romero <vromero at openjdk.org>
URL:       https://git.openjdk.java.net/valhalla/commit/b0450ed29b02eb84712632519ef28c0eabcc98b6
Stats:     71 lines in 5 files changed: 63 ins; 5 del; 3 mod

clean up to warnings related code

-------------

PR: https://git.openjdk.java.net/valhalla/pull/655

From dsimms at openjdk.java.net  Mon Feb 21 07:50:08 2022
From: dsimms at openjdk.java.net (David Simms)
Date: Mon, 21 Feb 2022 07:50:08 GMT
Subject: [lworld] RFR: 8281283: Remove usages of __WithField in runtime
 tests
In-Reply-To: <HyjRamI2444_LAuhOYTpn4wHfNCtVLxvoxG0kdAIx-U=.a88a4b33-56e9-4734-950d-7ca21f4fd593@github.com>
References: <HyjRamI2444_LAuhOYTpn4wHfNCtVLxvoxG0kdAIx-U=.a88a4b33-56e9-4734-950d-7ca21f4fd593@github.com>
Message-ID: <QvZm3kUmZ_g6l00-6ndtntqrN5El2MElsgSqRZx7KZg=.d833bdaa-36f4-45cd-bdbe-38aa050927fa@github.com>

On Thu, 17 Feb 2022 00:56:50 GMT, Dan Smith <dlsmith at openjdk.org> wrote:

> Modifying Valhalla runtime tests to no longer use `__WithField` as a macro for the `withfield` instruction. Most uses can be replaced with standard constructor compilation via javac. The remaining uses are rewritten in jasm.
> 
> Summary of changes:
> 
> - Remove all uses of `-XDallowWithFieldOperator`
> - Replace various `make` and `create` factory methods for primitive classes with constructors
> - Bug fixes for `JumboInline` that were exposed by this change
> - Rewrote `TestFieldNullability` to use jasm
> - Rewrote `VWithFieldTest` to use jasm, renamed `WithFieldTest`
> - Rewrote `WithFieldAccessorTest` to use jasm, made more comprehensive
> - Deleted `WithFieldNoAccessTest`, made redundant by `WithFieldAccessorTest` changes
> - Minor tweaks to `TestFieldTypeMismatch` for consistent jasm usage
> - Cleanup of some debugging output in `asmtools.JtregDriver`

For my part the Hotspot Runtime tests look good. Thanks for cleaning some of the old cruft where appropriate.

-------------

Marked as reviewed by dsimms (Committer).

PR: https://git.openjdk.java.net/valhalla/pull/653

From dsimms at openjdk.java.net  Mon Feb 21 07:50:13 2022
From: dsimms at openjdk.java.net (David Simms)
Date: Mon, 21 Feb 2022 07:50:13 GMT
Subject: [lworld] RFR: 8281283: Remove usages of __WithField in runtime
 tests
In-Reply-To: <sU9NZuE3vLdw8FgqP2rAA3ehciLVt7BVbCUJIv6Cns8=.b1965094-f587-49f6-912f-0d71b1e755ca@github.com>
References: <HyjRamI2444_LAuhOYTpn4wHfNCtVLxvoxG0kdAIx-U=.a88a4b33-56e9-4734-950d-7ca21f4fd593@github.com>
 <sU9NZuE3vLdw8FgqP2rAA3ehciLVt7BVbCUJIv6Cns8=.b1965094-f587-49f6-912f-0d71b1e755ca@github.com>
Message-ID: <zl_4hFeatcDBR15w2RbAihZeDly0XqwvA5xBZzRTa2E=.9c8bf66d-67d9-49b3-b077-e19e4ff44624@github.com>

On Thu, 17 Feb 2022 19:09:05 GMT, Dan Smith <dlsmith at openjdk.org> wrote:

>> Modifying Valhalla runtime tests to no longer use `__WithField` as a macro for the `withfield` instruction. Most uses can be replaced with standard constructor compilation via javac. The remaining uses are rewritten in jasm.
>> 
>> Summary of changes:
>> 
>> - Remove all uses of `-XDallowWithFieldOperator`
>> - Replace various `make` and `create` factory methods for primitive classes with constructors
>> - Bug fixes for `JumboInline` that were exposed by this change
>> - Rewrote `TestFieldNullability` to use jasm
>> - Rewrote `VWithFieldTest` to use jasm, renamed `WithFieldTest`
>> - Rewrote `WithFieldAccessorTest` to use jasm, made more comprehensive
>> - Deleted `WithFieldNoAccessTest`, made redundant by `WithFieldAccessorTest` changes
>> - Minor tweaks to `TestFieldTypeMismatch` for consistent jasm usage
>> - Cleanup of some debugging output in `asmtools.JtregDriver`
>
> test/hotspot/jtreg/runtime/valhalla/inlinetypes/JumboInline.java line 79:
> 
>> 77:         j = __WithField(j.l1, l1);
>> 78:         j = __WithField(j.l2, l0 + l1);
>> 79:         j = __WithField(j.l3, l1 + l2);
> 
> Bug here (and the following lines): `l2` refers to `this.l2`, not `j.l2`?if `this` is `JumboInline.default`, that value is 0, not `l0+l1`.

Agreed

> test/hotspot/jtreg/runtime/valhalla/inlinetypes/JumboInline.java line 113:
> 
>> 111:             return (l0 == j.l0 && l1 == j.l1 && l2 == j.l2 && l3 == j.l3
>> 112:                     && l4 == j.l4 && l5 == j.l5 && l6 == j.l6 && l7 == j.l7
>> 113:                     && l8 == j.l8 && l9 == j.l9 && l10 == j.l10 && l7 == j.l10
> 
> `l7 == j.l10`? Happens to work, because these are always 0, per the above bug.

The newer code here is cleaner, and more importantly correct. Again, agree, it only worked since "0" default

> test/hotspot/jtreg/runtime/valhalla/inlinetypes/TestFieldTypeMismatch.java line 32:
> 
>> 30:  * @library /test/lib
>> 31:  * @build org.openjdk.asmtools.* org.openjdk.asmtools.jasm.*
>> 32:  * @run driver org.openjdk.asmtools.JtregDriver jasm -strict TestFieldTypeMismatchClasses.jasm
> 
> Naming convention change: suggest putting jasm classes in a file named `[Test name]Classes.jasm`. No need for any classes declared there to actually use that name, although they should have names that won't clash with other tests in this directory.

Good idea

> test/hotspot/jtreg/runtime/valhalla/inlinetypes/TestValue4.java line 48:
> 
>> 46: 
>> 47:     public TestValue4() {
>> 48:         this((int) System.nanoTime());
> 
> Don't need the bit twiddling logic, because it's reproduced by the ByteBuffer reads in the other constructor.

At some point early on, debugging the interpreter was probably easy to not leave the frame, hence the manual bit twiddle, but I think we are past that now, nice clean up.

> test/hotspot/jtreg/runtime/valhalla/inlinetypes/WithFieldAccessorTest.java line 66:
> 
>> 64:         catchAccessError(() -> WithFieldSamePackage.withL(start, 10));
>> 65:         catchAccessError(() -> WithFieldSamePackage.withD(start, 11));
>> 66:         catchAccessError(() -> WithFieldSamePackage.withI(start, 12));
> 
> These calls replace the WithFieldNoAccessTest. (Not checking error messages. Should I add that?)

I think error message was more useful when things were unstable and this test broke more easily. Assume the frame information is correct and offending line number can be determined. Looks good.

-------------

PR: https://git.openjdk.java.net/valhalla/pull/653

From sadayapalam at openjdk.java.net  Tue Feb 22 09:39:57 2022
From: sadayapalam at openjdk.java.net (Srikanth Adayapalam)
Date: Tue, 22 Feb 2022 09:39:57 GMT
Subject: [lworld] RFR: 8281323: [lworld] Unnecessary entries in Preload
 attribute
Message-ID: <EptcSazUaXVaHDQ7T8DlWMLEse4NtP9a4esNpH54hGM=.728aebaa-f91a-43cc-b3df-69640971f57b@github.com>

Reimplement generation of Preload attribute decoupling it from InnerClasses attribute generation to minimize/eliminate needless entries from the Preload list

-------------

Commit messages:
 - Fix white space errors
 - 8281323:  Unnecessary entries in Preload attribute

Changes: https://git.openjdk.java.net/valhalla/pull/656/files
 Webrev: https://webrevs.openjdk.java.net/?repo=valhalla&pr=656&range=00
  Issue: https://bugs.openjdk.java.net/browse/JDK-8281323
  Stats: 342 lines in 8 files changed: 268 ins; 46 del; 28 mod
  Patch: https://git.openjdk.java.net/valhalla/pull/656.diff
  Fetch: git fetch https://git.openjdk.java.net/valhalla pull/656/head:pull/656

PR: https://git.openjdk.java.net/valhalla/pull/656

From sadayapalam at openjdk.java.net  Tue Feb 22 09:45:12 2022
From: sadayapalam at openjdk.java.net (Srikanth Adayapalam)
Date: Tue, 22 Feb 2022 09:45:12 GMT
Subject: [lworld] RFR: 8281323: [lworld] Unnecessary entries in Preload
 attribute
In-Reply-To: <EptcSazUaXVaHDQ7T8DlWMLEse4NtP9a4esNpH54hGM=.728aebaa-f91a-43cc-b3df-69640971f57b@github.com>
References: <EptcSazUaXVaHDQ7T8DlWMLEse4NtP9a4esNpH54hGM=.728aebaa-f91a-43cc-b3df-69640971f57b@github.com>
Message-ID: <JW2Ixqwg3Oh9SLzVk_4zH1kH8tC0slnvoVHGTpTxcRU=.8c94ff30-b919-4b33-a721-49b390f1be7e@github.com>

On Tue, 22 Feb 2022 09:29:37 GMT, Srikanth Adayapalam <sadayapalam at openjdk.org> wrote:

> Reimplement generation of Preload attribute decoupling it from InnerClasses attribute generation to minimize/eliminate needless entries from the Preload list

Notes for reviewer:

This change basically decouples the generation of Preload attribute from InnerClasses attribute - that scheme generates redundant entries as shows by https://bugs.openjdk.java.net/browse/JDK-8281323 and also as can be determined by a review of the draft JEP at https://openjdk.java.net/jeps/8277163 which simply says:


Each class file generated by javac includes a Preload attribute naming any value class that appears in one of the class file's declared field or method descriptors.


In internal discussions, the VM team also has expressed an interest in seeing those value classes that appear as types of local variables in Preload list. 

In Internal discussions, it was pointed out that where arrays are involved, if the component type happens to be a value type, it is not useful to see a Preload entry for that value class.

This PR also happens to solve https://bugs.openjdk.java.net/browse/JDK-8280942 - the reimplementation simply subsumes that fix too.

-------------

PR: https://git.openjdk.java.net/valhalla/pull/656

From scott at adligo.com  Tue Feb 22 01:44:51 2022
From: scott at adligo.com (Scott Morgan)
Date: Mon, 21 Feb 2022 19:44:51 -0600
Subject: Fwd: Custom Primitives?
In-Reply-To: <CANEdHmg4_R+kT0G0-DKTMjx+sJSgzXrzu9YPifNkCUnyLK=-XQ@mail.gmail.com>
References: <CANEdHmg4_R+kT0G0-DKTMjx+sJSgzXrzu9YPifNkCUnyLK=-XQ@mail.gmail.com>
Message-ID: <CANEdHmhshSUg2evhMirM5bze8=2uLTG3wK0HiVccMVu0ZNcBjw@mail.gmail.com>

Hi All,

   I've been messing around with some data structures / arrays lately and I
am wondering if Java will ever have custom primitives, and if those
custom primitives could contain optional pointers in some way.  I have
found a number of similar things Value Classes,
https://openjdk.java.net/jeps/401
Primitive Value Classes, etc, but I'm really looking for something a little
lower level.

Ideally I would like to store an array (with a single pointer pointing at
it) of one boolean and one optional pointer / reference, in a primitive
array structure to reduce the space of the null pointers.
   However this creates some obvious problems as length of the boolean only
could be one bit and the boolean + the pointer / reference could be 8 bytes
and one bit (on 64 bit JVMs).  Without the fixed length size of the
primitive, I imagine that the mapping to bytes and array indexes would be
quite challenging & slow.  However I have a gut feeling that something like
this path must be possible.

Are there any plans to solve problems like these in the current JEPs?
Is anyone doing similar research?

Cheers,
-- 
Regards,
Scott Morgan
President & CEO
Adligo Inc
http://www.adligo.com
https://www.linkedin.com/in/scott-morgan-21739415
A+ Better Business Bureau Rating
<https://www.bbb.org/chicago/business-reviews/computer-software-publishers-and-developers/adligo-inc-in-chicago-il-88381256>
https://github.com/adligo

By Appointment Only:
1-866-968-1893 Ex 101
scott at adligo.com
skype:adligo1?call
Send Me Files Securely:
*https://www.sendthisfile.com/f.jsp?id=ewOnyeFQM18IDRf7MMIdolfI
<https://www.sendthisfile.com/f.jsp?id=ewOnyeFQM18IDRf7MMIdolfI>*
https://discord.com/
Adligo#3066





-- 
Regards,
Scott Morgan
President & CEO
Adligo Inc
http://www.adligo.com
https://www.linkedin.com/in/scott-morgan-21739415
A+ Better Business Bureau Rating
<https://www.bbb.org/chicago/business-reviews/computer-software-publishers-and-developers/adligo-inc-in-chicago-il-88381256>
https://github.com/adligo

By Appointment Only:
1-866-968-1893 Ex 101
scott at adligo.com
skype:adligo1?call
Send Me Files Securely:
*https://www.sendthisfile.com/f.jsp?id=ewOnyeFQM18IDRf7MMIdolfI
<https://www.sendthisfile.com/f.jsp?id=ewOnyeFQM18IDRf7MMIdolfI>*
https://discord.com/
Adligo#3066

From sadayapalam at openjdk.java.net  Wed Feb 23 09:13:34 2022
From: sadayapalam at openjdk.java.net (Srikanth Adayapalam)
Date: Wed, 23 Feb 2022 09:13:34 GMT
Subject: [lworld] Integrated: 8282108: [lworld] Enhance CreateSymbols to read
 Preload attribute
In-Reply-To: <_211zSDzXdBGVEWhq25rnLc6p51RH8fm_RozwYlJW_w=.2d523564-15e3-48a6-ab01-b3d327899ffd@github.com>
References: <_211zSDzXdBGVEWhq25rnLc6p51RH8fm_RozwYlJW_w=.2d523564-15e3-48a6-ab01-b3d327899ffd@github.com>
Message-ID: <qzsMSq1Sc_i3GzrYaxF9nUOLN3mGbfQsoc-xv0hTkIE=.eed5d951-f7f6-498a-a4a9-a313f04c19f0@github.com>

On Wed, 23 Feb 2022 09:06:34 GMT, Srikanth Adayapalam <sadayapalam at openjdk.org> wrote:

> Arrange to skip Preload attribute as it is not relevant to compilation

This pull request has now been integrated.

Changeset: 52546456
Author:    Srikanth Adayapalam <sadayapalam at openjdk.org>
URL:       https://git.openjdk.java.net/valhalla/commit/52546456dd9c53152f807a1306a2ff3e19cfab2e
Stats:     1 line in 1 file changed: 1 ins; 0 del; 0 mod

8282108: [lworld] Enhance CreateSymbols to read Preload attribute

-------------

PR: https://git.openjdk.java.net/valhalla/pull/657

From sadayapalam at openjdk.java.net  Wed Feb 23 09:13:33 2022
From: sadayapalam at openjdk.java.net (Srikanth Adayapalam)
Date: Wed, 23 Feb 2022 09:13:33 GMT
Subject: [lworld] Integrated: 8282108: [lworld] Enhance CreateSymbols to read
 Preload attribute
Message-ID: <_211zSDzXdBGVEWhq25rnLc6p51RH8fm_RozwYlJW_w=.2d523564-15e3-48a6-ab01-b3d327899ffd@github.com>

Arrange to skip Preload attribute as it is not relevant to compilation

-------------

Commit messages:
 - 8282108: Enhance CreateSymbols to read Preload attribute

Changes: https://git.openjdk.java.net/valhalla/pull/657/files
 Webrev: https://webrevs.openjdk.java.net/?repo=valhalla&pr=657&range=00
  Issue: https://bugs.openjdk.java.net/browse/JDK-8282108
  Stats: 1 line in 1 file changed: 1 ins; 0 del; 0 mod
  Patch: https://git.openjdk.java.net/valhalla/pull/657.diff
  Fetch: git fetch https://git.openjdk.java.net/valhalla pull/657/head:pull/657

PR: https://git.openjdk.java.net/valhalla/pull/657

From duke at openjdk.java.net  Wed Feb 23 10:39:21 2022
From: duke at openjdk.java.net (Aggelos Biboudis)
Date: Wed, 23 Feb 2022 10:39:21 GMT
Subject: [lworld] RFR: 8281323: [lworld] Unnecessary entries in Preload
 attribute
In-Reply-To: <EptcSazUaXVaHDQ7T8DlWMLEse4NtP9a4esNpH54hGM=.728aebaa-f91a-43cc-b3df-69640971f57b@github.com>
References: <EptcSazUaXVaHDQ7T8DlWMLEse4NtP9a4esNpH54hGM=.728aebaa-f91a-43cc-b3df-69640971f57b@github.com>
Message-ID: <vC3sCsQdnZzg0SbO4O2f66N2rl_J98xCMONrbeasL4Y=.1bd9b3d3-b660-4962-9914-73461c082c71@github.com>

On Tue, 22 Feb 2022 09:29:37 GMT, Srikanth Adayapalam <sadayapalam at openjdk.org> wrote:

> Reimplement generation of Preload attribute decoupling it from InnerClasses attribute generation to minimize/eliminate needless entries from the Preload list

Looks good!

src/jdk.compiler/share/classes/com/sun/tools/javac/jvm/PoolWriter.java line 260:

> 258:         }
> 259:         c.complete();
> 260:         if (c.isValueClass() && !c.isPrimitiveClass()) {

This is fully covered by `requiresPreload` in `Type`. Actually more strict, because of the `if (this.isReferenceProjection()) return true;` condition in the previous method. Is my understanding correct?

-------------

Marked as reviewed by biboudis at github.com (no known OpenJDK username).

PR: https://git.openjdk.java.net/valhalla/pull/656

From sadayapalam at openjdk.java.net  Wed Feb 23 10:54:17 2022
From: sadayapalam at openjdk.java.net (Srikanth Adayapalam)
Date: Wed, 23 Feb 2022 10:54:17 GMT
Subject: [lworld] RFR: 8281323: [lworld] Unnecessary entries in Preload
 attribute
In-Reply-To: <vC3sCsQdnZzg0SbO4O2f66N2rl_J98xCMONrbeasL4Y=.1bd9b3d3-b660-4962-9914-73461c082c71@github.com>
References: <EptcSazUaXVaHDQ7T8DlWMLEse4NtP9a4esNpH54hGM=.728aebaa-f91a-43cc-b3df-69640971f57b@github.com>
 <vC3sCsQdnZzg0SbO4O2f66N2rl_J98xCMONrbeasL4Y=.1bd9b3d3-b660-4962-9914-73461c082c71@github.com>
Message-ID: <vDGznP_0t4qTNjXc2nEdQCoa_9NjmF7nywq2uGvGxMU=.c66f8317-fdc0-492f-aab5-bc320ed6d69a@github.com>

On Wed, 23 Feb 2022 10:33:30 GMT, Aggelos Biboudis <duke at openjdk.java.net> wrote:

> This is fully covered by `requiresPreload` in `Type`. Actually more strict, because of the `if (this.isReferenceProjection()) return true;` condition in the previous method. Is my understanding correct?

Yes, you are correct. The check

if (this.isReferenceProjection()) return true;

covers the requirement for https://bugs.openjdk.java.net/browse/JDK-8280942

The check

return this.isValueClass() && !this.isPrimitiveClass();

works for "pure" B2 classes. (every B3 is also a B2)

So we are saying in effect preload is required for pure B2 classes or for a B3 class whose reference projection is used (in field/method descriptors and/or local types)

-------------

PR: https://git.openjdk.java.net/valhalla/pull/656

From sadayapalam at openjdk.java.net  Wed Feb 23 10:57:24 2022
From: sadayapalam at openjdk.java.net (Srikanth Adayapalam)
Date: Wed, 23 Feb 2022 10:57:24 GMT
Subject: [lworld] Integrated: 8281323: [lworld] Unnecessary entries in Preload
 attribute
In-Reply-To: <EptcSazUaXVaHDQ7T8DlWMLEse4NtP9a4esNpH54hGM=.728aebaa-f91a-43cc-b3df-69640971f57b@github.com>
References: <EptcSazUaXVaHDQ7T8DlWMLEse4NtP9a4esNpH54hGM=.728aebaa-f91a-43cc-b3df-69640971f57b@github.com>
Message-ID: <4cib0jjnrs7wfw8brK2dj5zs5KHNAlUyQJ0mZ2_O9c0=.83a1e8e3-b433-445c-8f0d-156f22e7afd8@github.com>

On Tue, 22 Feb 2022 09:29:37 GMT, Srikanth Adayapalam <sadayapalam at openjdk.org> wrote:

> Reimplement generation of Preload attribute decoupling it from InnerClasses attribute generation to minimize/eliminate needless entries from the Preload list

This pull request has now been integrated.

Changeset: 603cb9d6
Author:    Srikanth Adayapalam <sadayapalam at openjdk.org>
URL:       https://git.openjdk.java.net/valhalla/commit/603cb9d66be577ad17f90a0386160761d4a76b06
Stats:     342 lines in 8 files changed: 268 ins; 46 del; 28 mod

8281323: [lworld] Unnecessary entries in Preload attribute
8280942: Preload attribute should mention primitive classes when reference projection is used in descriptors

-------------

PR: https://git.openjdk.java.net/valhalla/pull/656

From vromero at openjdk.java.net  Wed Feb 23 20:54:47 2022
From: vromero at openjdk.java.net (Vicente Romero)
Date: Wed, 23 Feb 2022 20:54:47 GMT
Subject: RFR: type of TypeVar.ref is incorrect and should be fixed
Message-ID: <3vMo5X6OcSNOAgd-VY3YTyTyx5eEUbDQ-ARZGQBtyVA=.4ed18c59-7fbb-43a9-9464-48a3b636ca78@github.com>

The type of TypeVar.ref is incorrect and should be fixed, in particular due to a bug, the compiler is not able to differentiate between  `T` and `T.ref`. Also did some cleanup to warnings related code, in particular there should be a different warning for cases when a universal tvar is not assigned `null` but it is assigned a value that could be null at run time. This warning should not be the same as the one for a plain null assignment.

-------------

Commit messages:
 - clean up to warnings related code

Changes: https://git.openjdk.java.net/valhalla/pull/658/files
 Webrev: https://webrevs.openjdk.java.net/?repo=valhalla&pr=658&range=00
  Stats: 87 lines in 6 files changed: 77 ins; 5 del; 5 mod
  Patch: https://git.openjdk.java.net/valhalla/pull/658.diff
  Fetch: git fetch https://git.openjdk.java.net/valhalla pull/658/head:pull/658

PR: https://git.openjdk.java.net/valhalla/pull/658

From dlsmith at openjdk.java.net  Wed Feb 23 23:33:28 2022
From: dlsmith at openjdk.java.net (Dan Smith)
Date: Wed, 23 Feb 2022 23:33:28 GMT
Subject: git: openjdk/valhalla: lworld: 8281283: Remove usages of __WithField
 in runtime tests
Message-ID: <3c052f64-8971-4173-8d22-fcb0a6cd5d57@openjdk.org>

Changeset: 71cf8628
Author:    Dan Smith <dlsmith at openjdk.org>
Date:      2022-02-23 23:32:57 +0000
URL:       https://git.openjdk.java.net/valhalla/commit/71cf8628fa3133ff43bff0238f7a4eaa921cb88d

8281283: Remove usages of __WithField in runtime tests

Reviewed-by: rriggs, dsimms

! test/hotspot/jtreg/runtime/valhalla/inlinetypes/ClassPrintLayoutDcmd.java
! test/hotspot/jtreg/runtime/valhalla/inlinetypes/FlattenableSemanticTest.java
! test/hotspot/jtreg/runtime/valhalla/inlinetypes/InlineOops.java
! test/hotspot/jtreg/runtime/valhalla/inlinetypes/InlineTypeArray.java
! test/hotspot/jtreg/runtime/valhalla/inlinetypes/InlineTypeCreation.java
! test/hotspot/jtreg/runtime/valhalla/inlinetypes/InlineTypeDensity.java
! test/hotspot/jtreg/runtime/valhalla/inlinetypes/InlineTypeGetField.java
! test/hotspot/jtreg/runtime/valhalla/inlinetypes/InlineTypesTest.java
! test/hotspot/jtreg/runtime/valhalla/inlinetypes/InlineWithJni.java
! test/hotspot/jtreg/runtime/valhalla/inlinetypes/JumboInline.java
! test/hotspot/jtreg/runtime/valhalla/inlinetypes/Long8Inline.java
! test/hotspot/jtreg/runtime/valhalla/inlinetypes/ObjectMethods.java
! test/hotspot/jtreg/runtime/valhalla/inlinetypes/Person.java
! test/hotspot/jtreg/runtime/valhalla/inlinetypes/Point.java
! test/hotspot/jtreg/runtime/valhalla/inlinetypes/QuickeningTest.java
! test/hotspot/jtreg/runtime/valhalla/inlinetypes/Test8186715.java
! test/hotspot/jtreg/runtime/valhalla/inlinetypes/TestFieldNullability.java
+ test/hotspot/jtreg/runtime/valhalla/inlinetypes/TestFieldNullabilityClasses.jasm
! test/hotspot/jtreg/runtime/valhalla/inlinetypes/TestFieldTypeMismatch.java
+ test/hotspot/jtreg/runtime/valhalla/inlinetypes/TestFieldTypeMismatchClasses.jasm
- test/hotspot/jtreg/runtime/valhalla/inlinetypes/TestFieldTypeMismatchHelper.jasm
! test/hotspot/jtreg/runtime/valhalla/inlinetypes/TestInheritedInlineTypeFields.java
! test/hotspot/jtreg/runtime/valhalla/inlinetypes/TestValue1.java
! test/hotspot/jtreg/runtime/valhalla/inlinetypes/TestValue2.java
! test/hotspot/jtreg/runtime/valhalla/inlinetypes/TestValue3.java
! test/hotspot/jtreg/runtime/valhalla/inlinetypes/TestValue4.java
! test/hotspot/jtreg/runtime/valhalla/inlinetypes/UninitializedInlineFieldsTest.java
! test/hotspot/jtreg/runtime/valhalla/inlinetypes/UnsafeTest.java
! test/hotspot/jtreg/runtime/valhalla/inlinetypes/VDefaultTest.java
- test/hotspot/jtreg/runtime/valhalla/inlinetypes/VWithFieldTest.java
! test/hotspot/jtreg/runtime/valhalla/inlinetypes/WithFieldAccessorTest.java
+ test/hotspot/jtreg/runtime/valhalla/inlinetypes/WithFieldAccessorTestClasses.jasm
- test/hotspot/jtreg/runtime/valhalla/inlinetypes/WithFieldNoAccessTest.jcod
+ test/hotspot/jtreg/runtime/valhalla/inlinetypes/WithFieldTest.java
+ test/hotspot/jtreg/runtime/valhalla/inlinetypes/WithFieldTestClasses.jasm
! test/hotspot/jtreg/runtime/valhalla/inlinetypes/verifier/VTAssignability.java
! test/hotspot/jtreg/runtime/valhalla/inlinetypes/verifier/VTMonitor.java
! test/lib/org/openjdk/asmtools/JtregDriver.java


From dlsmith at openjdk.java.net  Wed Feb 23 23:36:29 2022
From: dlsmith at openjdk.java.net (Dan Smith)
Date: Wed, 23 Feb 2022 23:36:29 GMT
Subject: [lworld] Integrated: 8281283: Remove usages of __WithField in runtime
 tests
In-Reply-To: <HyjRamI2444_LAuhOYTpn4wHfNCtVLxvoxG0kdAIx-U=.a88a4b33-56e9-4734-950d-7ca21f4fd593@github.com>
References: <HyjRamI2444_LAuhOYTpn4wHfNCtVLxvoxG0kdAIx-U=.a88a4b33-56e9-4734-950d-7ca21f4fd593@github.com>
Message-ID: <loXUCVeinTMyhWWpHnDCyyg_iLs7Vce-7N8gNMicF14=.11a8e351-fb03-4031-87e9-ccd367e1a05f@github.com>

On Thu, 17 Feb 2022 00:56:50 GMT, Dan Smith <dlsmith at openjdk.org> wrote:

> Modifying Valhalla runtime tests to no longer use `__WithField` as a macro for the `withfield` instruction. Most uses can be replaced with standard constructor compilation via javac. The remaining uses are rewritten in jasm.
> 
> Summary of changes:
> 
> - Remove all uses of `-XDallowWithFieldOperator`
> - Replace various `make` and `create` factory methods for primitive classes with constructors
> - Bug fixes for `JumboInline` that were exposed by this change
> - Rewrote `TestFieldNullability` to use jasm
> - Rewrote `VWithFieldTest` to use jasm, renamed `WithFieldTest`
> - Rewrote `WithFieldAccessorTest` to use jasm, made more comprehensive
> - Deleted `WithFieldNoAccessTest`, made redundant by `WithFieldAccessorTest` changes
> - Minor tweaks to `TestFieldTypeMismatch` for consistent jasm usage
> - Cleanup of some debugging output in `asmtools.JtregDriver`

This pull request has now been integrated.

Changeset: 71cf8628
Author:    Dan Smith <dlsmith at openjdk.org>
URL:       https://git.openjdk.java.net/valhalla/commit/71cf8628fa3133ff43bff0238f7a4eaa921cb88d
Stats:     2016 lines in 38 files changed: 703 ins; 1105 del; 208 mod

8281283: Remove usages of __WithField in runtime tests

Reviewed-by: rriggs, dsimms

-------------

PR: https://git.openjdk.java.net/valhalla/pull/653

From vromero at openjdk.java.net  Thu Feb 24 18:25:13 2022
From: vromero at openjdk.java.net (Vicente Romero)
Date: Thu, 24 Feb 2022 18:25:13 GMT
Subject: RFR: type of TypeVar.ref is incorrect and should be fixed [v2]
In-Reply-To: <3vMo5X6OcSNOAgd-VY3YTyTyx5eEUbDQ-ARZGQBtyVA=.4ed18c59-7fbb-43a9-9464-48a3b636ca78@github.com>
References: <3vMo5X6OcSNOAgd-VY3YTyTyx5eEUbDQ-ARZGQBtyVA=.4ed18c59-7fbb-43a9-9464-48a3b636ca78@github.com>
Message-ID: <jrScphd4PMKvslQ4VKwIaZ9WgnO_qcv3jY1Ey-b28ak=.bd235ffe-de5a-46a6-8ed2-e24031ddef71@github.com>

> The type of TypeVar.ref is incorrect and should be fixed, in particular due to a bug, the compiler is not able to differentiate between  `T` and `T.ref`. Also did some cleanup to warnings related code, in particular there should be a different warning for cases when a universal tvar is not assigned `null` but it is assigned a value that could be null at run time. This warning should not be the same as the one for a plain null assignment.

Vicente Romero has updated the pull request with a new target base due to a merge or a rebase. The incremental webrev excludes the unrelated changes brought in by the merge/rebase. The pull request contains two additional commits since the last revision:

 - Merge branch 'universal-tvars' into fixing.warning.bugs
 - clean up to warnings related code

-------------

Changes:
  - all: https://git.openjdk.java.net/valhalla/pull/658/files
  - new: https://git.openjdk.java.net/valhalla/pull/658/files/b3edd868..c8bb952a

Webrevs:
 - full: https://webrevs.openjdk.java.net/?repo=valhalla&pr=658&range=01
 - incr: https://webrevs.openjdk.java.net/?repo=valhalla&pr=658&range=00-01

  Stats: 0 lines in 0 files changed: 0 ins; 0 del; 0 mod
  Patch: https://git.openjdk.java.net/valhalla/pull/658.diff
  Fetch: git fetch https://git.openjdk.java.net/valhalla pull/658/head:pull/658

PR: https://git.openjdk.java.net/valhalla/pull/658

From vromero at openjdk.java.net  Fri Feb 25 00:29:09 2022
From: vromero at openjdk.java.net (Vicente Romero)
Date: Fri, 25 Feb 2022 00:29:09 GMT
Subject: RFR: type of TypeVar.ref is incorrect and should be fixed [v3]
In-Reply-To: <3vMo5X6OcSNOAgd-VY3YTyTyx5eEUbDQ-ARZGQBtyVA=.4ed18c59-7fbb-43a9-9464-48a3b636ca78@github.com>
References: <3vMo5X6OcSNOAgd-VY3YTyTyx5eEUbDQ-ARZGQBtyVA=.4ed18c59-7fbb-43a9-9464-48a3b636ca78@github.com>
Message-ID: <eBOUvAM92QR5CIKxFQJygJ4e0eGC00rNHc-DNOiANsw=.6374b9b7-30ca-44bd-8fd7-b0e2373fdc21@github.com>

> The type of TypeVar.ref is incorrect and should be fixed, in particular due to a bug, the compiler is not able to differentiate between  `T` and `T.ref` which should have a different type

Vicente Romero has updated the pull request incrementally with one additional commit since the last revision:

  more tests

-------------

Changes:
  - all: https://git.openjdk.java.net/valhalla/pull/658/files
  - new: https://git.openjdk.java.net/valhalla/pull/658/files/c8bb952a..50f7ae2f

Webrevs:
 - full: https://webrevs.openjdk.java.net/?repo=valhalla&pr=658&range=02
 - incr: https://webrevs.openjdk.java.net/?repo=valhalla&pr=658&range=01-02

  Stats: 107 lines in 1 file changed: 53 ins; 26 del; 28 mod
  Patch: https://git.openjdk.java.net/valhalla/pull/658.diff
  Fetch: git fetch https://git.openjdk.java.net/valhalla pull/658/head:pull/658

PR: https://git.openjdk.java.net/valhalla/pull/658

From vromero at openjdk.java.net  Fri Feb 25 01:29:43 2022
From: vromero at openjdk.java.net (Vicente Romero)
Date: Fri, 25 Feb 2022 01:29:43 GMT
Subject: RFR: type of TypeVar.ref is incorrect and should be fixed [v4]
In-Reply-To: <3vMo5X6OcSNOAgd-VY3YTyTyx5eEUbDQ-ARZGQBtyVA=.4ed18c59-7fbb-43a9-9464-48a3b636ca78@github.com>
References: <3vMo5X6OcSNOAgd-VY3YTyTyx5eEUbDQ-ARZGQBtyVA=.4ed18c59-7fbb-43a9-9464-48a3b636ca78@github.com>
Message-ID: <WOBslVGZvs17ij4bfHQQ03S33NKp-c7OztQhbqiJZC8=.c542311d-985b-4a67-9b55-5a34f77213ad@github.com>

> The type of TypeVar.ref is incorrect and should be fixed, in particular due to a bug, the compiler is not able to differentiate between  `T` and `T.ref` which should have a different type

Vicente Romero has updated the pull request incrementally with one additional commit since the last revision:

  minor refactoring

-------------

Changes:
  - all: https://git.openjdk.java.net/valhalla/pull/658/files
  - new: https://git.openjdk.java.net/valhalla/pull/658/files/50f7ae2f..df161eb5

Webrevs:
 - full: https://webrevs.openjdk.java.net/?repo=valhalla&pr=658&range=03
 - incr: https://webrevs.openjdk.java.net/?repo=valhalla&pr=658&range=02-03

  Stats: 8 lines in 1 file changed: 5 ins; 0 del; 3 mod
  Patch: https://git.openjdk.java.net/valhalla/pull/658.diff
  Fetch: git fetch https://git.openjdk.java.net/valhalla pull/658/head:pull/658

PR: https://git.openjdk.java.net/valhalla/pull/658

From vromero at openjdk.java.net  Fri Feb 25 01:35:55 2022
From: vromero at openjdk.java.net (Vicente Romero)
Date: Fri, 25 Feb 2022 01:35:55 GMT
Subject: RFR: type of TypeVar.ref is incorrect and should be fixed [v5]
In-Reply-To: <3vMo5X6OcSNOAgd-VY3YTyTyx5eEUbDQ-ARZGQBtyVA=.4ed18c59-7fbb-43a9-9464-48a3b636ca78@github.com>
References: <3vMo5X6OcSNOAgd-VY3YTyTyx5eEUbDQ-ARZGQBtyVA=.4ed18c59-7fbb-43a9-9464-48a3b636ca78@github.com>
Message-ID: <kFUOAxtr620BBOfJ3wQITDgxxCSX1SigF1ykfpPMQEk=.f8afba92-6cb2-4aee-b268-1c8a15881a22@github.com>

> The type of TypeVar.ref is incorrect and should be fixed, in particular due to a bug, the compiler is not able to differentiate between  `T` and `T.ref` which should have a different type

Vicente Romero has updated the pull request incrementally with one additional commit since the last revision:

  removing whitespace

-------------

Changes:
  - all: https://git.openjdk.java.net/valhalla/pull/658/files
  - new: https://git.openjdk.java.net/valhalla/pull/658/files/df161eb5..4d9379db

Webrevs:
 - full: https://webrevs.openjdk.java.net/?repo=valhalla&pr=658&range=04
 - incr: https://webrevs.openjdk.java.net/?repo=valhalla&pr=658&range=03-04

  Stats: 1 line in 1 file changed: 0 ins; 0 del; 1 mod
  Patch: https://git.openjdk.java.net/valhalla/pull/658.diff
  Fetch: git fetch https://git.openjdk.java.net/valhalla pull/658/head:pull/658

PR: https://git.openjdk.java.net/valhalla/pull/658

From fparain at openjdk.java.net  Fri Feb 25 14:49:03 2022
From: fparain at openjdk.java.net (Frederic Parain)
Date: Fri, 25 Feb 2022 14:49:03 GMT
Subject: [lworld] RFR: 8282292: [lworld]
 runtime/valhalla/inlinetypes/WithFieldAccessorTest.java Xcomp C1 asserts
Message-ID: <zubP2hmyEJU-2pNyCFulTqD9DK6yxCqvWKhEj2HQ434=.5d132f5f-4634-45fc-8b0b-d8db2f8273c2@github.com>

Please review this small patch fixing handling of illegal accesses in C1's implementation of withfield.

Thank you,

Fred

-------------

Commit messages:
 - Withfield: insert deopt point if field is inaccessible

Changes: https://git.openjdk.java.net/valhalla/pull/659/files
 Webrev: https://webrevs.openjdk.java.net/?repo=valhalla&pr=659&range=00
  Issue: https://bugs.openjdk.java.net/browse/JDK-8282292
  Stats: 1 line in 1 file changed: 0 ins; 0 del; 1 mod
  Patch: https://git.openjdk.java.net/valhalla/pull/659.diff
  Fetch: git fetch https://git.openjdk.java.net/valhalla pull/659/head:pull/659

PR: https://git.openjdk.java.net/valhalla/pull/659

From fparain at openjdk.java.net  Fri Feb 25 15:15:23 2022
From: fparain at openjdk.java.net (Frederic Parain)
Date: Fri, 25 Feb 2022 15:15:23 GMT
Subject: git: openjdk/valhalla: lworld: 8282292: [lworld]
 runtime/valhalla/inlinetypes/WithFieldAccessorTest.java Xcomp C1 asserts
Message-ID: <5667969d-eaa0-46f6-a484-957ea8a94e67@openjdk.org>

Changeset: 53bb0965
Author:    Frederic Parain <fparain at openjdk.org>
Date:      2022-02-25 15:14:58 +0000
URL:       https://git.openjdk.java.net/valhalla/commit/53bb09651a02a21b7cb2d18e9bcbd74e6198355d

8282292: [lworld] runtime/valhalla/inlinetypes/WithFieldAccessorTest.java Xcomp C1 asserts

Reviewed-by: thartmann

! src/hotspot/share/c1/c1_GraphBuilder.cpp


From fparain at openjdk.java.net  Fri Feb 25 15:18:21 2022
From: fparain at openjdk.java.net (Frederic Parain)
Date: Fri, 25 Feb 2022 15:18:21 GMT
Subject: [lworld] RFR: 8282292: [lworld]
 runtime/valhalla/inlinetypes/WithFieldAccessorTest.java Xcomp C1 asserts
In-Reply-To: <zubP2hmyEJU-2pNyCFulTqD9DK6yxCqvWKhEj2HQ434=.5d132f5f-4634-45fc-8b0b-d8db2f8273c2@github.com>
References: <zubP2hmyEJU-2pNyCFulTqD9DK6yxCqvWKhEj2HQ434=.5d132f5f-4634-45fc-8b0b-d8db2f8273c2@github.com>
Message-ID: <cJU-Ry858sWo6Ib1k5KBCIqKiniyjdxgpfV-nZYDRxc=.74aca0db-ddfc-47e5-9912-3e43785b6ee5@github.com>

On Fri, 25 Feb 2022 14:41:46 GMT, Frederic Parain <fparain at openjdk.org> wrote:

> Please review this small patch fixing handling of illegal accesses in C1's implementation of withfield.
> 
> Thank you,
> 
> Fred

Thanks for the review.

-------------

PR: https://git.openjdk.java.net/valhalla/pull/659

From thartmann at openjdk.java.net  Fri Feb 25 15:18:20 2022
From: thartmann at openjdk.java.net (Tobias Hartmann)
Date: Fri, 25 Feb 2022 15:18:20 GMT
Subject: [lworld] RFR: 8282292: [lworld]
 runtime/valhalla/inlinetypes/WithFieldAccessorTest.java Xcomp C1 asserts
In-Reply-To: <zubP2hmyEJU-2pNyCFulTqD9DK6yxCqvWKhEj2HQ434=.5d132f5f-4634-45fc-8b0b-d8db2f8273c2@github.com>
References: <zubP2hmyEJU-2pNyCFulTqD9DK6yxCqvWKhEj2HQ434=.5d132f5f-4634-45fc-8b0b-d8db2f8273c2@github.com>
Message-ID: <sjnAR3f-VcEsu9Z8gVbXUpeWpfm_I2i5leg84pOAHSE=.c4848cd1-a594-4fe1-b227-9a3944e2217e@github.com>

On Fri, 25 Feb 2022 14:41:46 GMT, Frederic Parain <fparain at openjdk.org> wrote:

> Please review this small patch fixing handling of illegal accesses in C1's implementation of withfield.
> 
> Thank you,
> 
> Fred

Looks good to me, thanks for fixing!

-------------

Marked as reviewed by thartmann (Committer).

PR: https://git.openjdk.java.net/valhalla/pull/659

From fparain at openjdk.java.net  Fri Feb 25 15:18:22 2022
From: fparain at openjdk.java.net (Frederic Parain)
Date: Fri, 25 Feb 2022 15:18:22 GMT
Subject: [lworld] Integrated: 8282292: [lworld]
 runtime/valhalla/inlinetypes/WithFieldAccessorTest.java Xcomp C1 asserts
In-Reply-To: <zubP2hmyEJU-2pNyCFulTqD9DK6yxCqvWKhEj2HQ434=.5d132f5f-4634-45fc-8b0b-d8db2f8273c2@github.com>
References: <zubP2hmyEJU-2pNyCFulTqD9DK6yxCqvWKhEj2HQ434=.5d132f5f-4634-45fc-8b0b-d8db2f8273c2@github.com>
Message-ID: <LkQJzvdHA2MO7OHpA78raCiQrFH9-rl0crOAHpGehiM=.94054833-5d97-4a97-a7d4-134ee9120007@github.com>

On Fri, 25 Feb 2022 14:41:46 GMT, Frederic Parain <fparain at openjdk.org> wrote:

> Please review this small patch fixing handling of illegal accesses in C1's implementation of withfield.
> 
> Thank you,
> 
> Fred

This pull request has now been integrated.

Changeset: 53bb0965
Author:    Frederic Parain <fparain at openjdk.org>
URL:       https://git.openjdk.java.net/valhalla/commit/53bb09651a02a21b7cb2d18e9bcbd74e6198355d
Stats:     1 line in 1 file changed: 0 ins; 0 del; 1 mod

8282292: [lworld] runtime/valhalla/inlinetypes/WithFieldAccessorTest.java Xcomp C1 asserts

Reviewed-by: thartmann

-------------

PR: https://git.openjdk.java.net/valhalla/pull/659

From mcimadamore at openjdk.java.net  Fri Feb 25 16:48:14 2022
From: mcimadamore at openjdk.java.net (Maurizio Cimadamore)
Date: Fri, 25 Feb 2022 16:48:14 GMT
Subject: RFR: type of TypeVar.ref is incorrect and should be fixed [v5]
In-Reply-To: <kFUOAxtr620BBOfJ3wQITDgxxCSX1SigF1ykfpPMQEk=.f8afba92-6cb2-4aee-b268-1c8a15881a22@github.com>
References: <3vMo5X6OcSNOAgd-VY3YTyTyx5eEUbDQ-ARZGQBtyVA=.4ed18c59-7fbb-43a9-9464-48a3b636ca78@github.com>
 <kFUOAxtr620BBOfJ3wQITDgxxCSX1SigF1ykfpPMQEk=.f8afba92-6cb2-4aee-b268-1c8a15881a22@github.com>
Message-ID: <rElnYiAl2IMsDZK7OgJ6GIZw_hIkz9xKN-eFmMPbIMs=.9e5cec1a-cd48-4b18-9379-bbd02d7fa1a7@github.com>

On Fri, 25 Feb 2022 01:35:55 GMT, Vicente Romero <vromero at openjdk.org> wrote:

>> The type of TypeVar.ref is incorrect and should be fixed, in particular due to a bug, the compiler is not able to differentiate between  `T` and `T.ref` which should have a different type
>
> Vicente Romero has updated the pull request incrementally with one additional commit since the last revision:
> 
>   removing whitespace

Looks good!

-------------

Marked as reviewed by mcimadamore (Committer).

PR: https://git.openjdk.java.net/valhalla/pull/658

From vromero at openjdk.java.net  Fri Feb 25 17:50:29 2022
From: vromero at openjdk.java.net (Vicente Romero)
Date: Fri, 25 Feb 2022 17:50:29 GMT
Subject: RFR: type of TypeVar.ref is incorrect and should be fixed [v6]
In-Reply-To: <3vMo5X6OcSNOAgd-VY3YTyTyx5eEUbDQ-ARZGQBtyVA=.4ed18c59-7fbb-43a9-9464-48a3b636ca78@github.com>
References: <3vMo5X6OcSNOAgd-VY3YTyTyx5eEUbDQ-ARZGQBtyVA=.4ed18c59-7fbb-43a9-9464-48a3b636ca78@github.com>
Message-ID: <2KxPiarrLVgdyrmfuivZ5nuTUlA4yUIgN7T-8js-hf4=.b347ba20-76ea-4f64-980b-e28773f9295c@github.com>

> The type of TypeVar.ref is incorrect and should be fixed, in particular due to a bug, the compiler is not able to differentiate between  `T` and `T.ref` which should have a different type

Vicente Romero has updated the pull request incrementally with one additional commit since the last revision:

  removing whitespace

-------------

Changes:
  - all: https://git.openjdk.java.net/valhalla/pull/658/files
  - new: https://git.openjdk.java.net/valhalla/pull/658/files/4d9379db..2b4f4416

Webrevs:
 - full: https://webrevs.openjdk.java.net/?repo=valhalla&pr=658&range=05
 - incr: https://webrevs.openjdk.java.net/?repo=valhalla&pr=658&range=04-05

  Stats: 1 line in 1 file changed: 0 ins; 0 del; 1 mod
  Patch: https://git.openjdk.java.net/valhalla/pull/658.diff
  Fetch: git fetch https://git.openjdk.java.net/valhalla pull/658/head:pull/658

PR: https://git.openjdk.java.net/valhalla/pull/658

From vromero at openjdk.java.net  Fri Feb 25 18:03:15 2022
From: vromero at openjdk.java.net (Vicente Romero)
Date: Fri, 25 Feb 2022 18:03:15 GMT
Subject: Integrated: type of TypeVar.ref is incorrect and should be fixed
In-Reply-To: <3vMo5X6OcSNOAgd-VY3YTyTyx5eEUbDQ-ARZGQBtyVA=.4ed18c59-7fbb-43a9-9464-48a3b636ca78@github.com>
References: <3vMo5X6OcSNOAgd-VY3YTyTyx5eEUbDQ-ARZGQBtyVA=.4ed18c59-7fbb-43a9-9464-48a3b636ca78@github.com>
Message-ID: <3181KV-xpjcOz-TuJnzs9mTT1eClxxIqa-8myGle1ms=.ca24a8b8-9fb1-4627-ae23-fe3c38d87523@github.com>

On Wed, 23 Feb 2022 20:44:36 GMT, Vicente Romero <vromero at openjdk.org> wrote:

> The type of TypeVar.ref is incorrect and should be fixed, in particular due to a bug, the compiler is not able to differentiate between  `T` and `T.ref` which should have a different type

This pull request has now been integrated.

Changeset: 8841a648
Author:    Vicente Romero <vromero at openjdk.org>
URL:       https://git.openjdk.java.net/valhalla/commit/8841a6485dd1c382b2f0f43edc089405d3daac6b
Stats:     117 lines in 2 files changed: 66 ins; 20 del; 31 mod

type of TypeVar.ref is incorrect and should be fixed

Reviewed-by: mcimadamore

-------------

PR: https://git.openjdk.java.net/valhalla/pull/658

From vromero at openjdk.java.net  Fri Feb 25 18:19:42 2022
From: vromero at openjdk.java.net (Vicente Romero)
Date: Fri, 25 Feb 2022 18:19:42 GMT
Subject: Integrated: some test refactorings
Message-ID: <wVR7mK39PxqsVRdeqBwIhqs1nwrus4aDIpDTD9d2V48=.fee5a766-bd7d-465c-a799-6586101ad698@github.com>

Refactoring tests at UniversalTvarsCompilationTests

-------------

Commit messages:
 - test refactorings

Changes: https://git.openjdk.java.net/valhalla/pull/660/files
 Webrev: https://webrevs.openjdk.java.net/?repo=valhalla&pr=660&range=00
  Stats: 69 lines in 1 file changed: 8 ins; 39 del; 22 mod
  Patch: https://git.openjdk.java.net/valhalla/pull/660.diff
  Fetch: git fetch https://git.openjdk.java.net/valhalla pull/660/head:pull/660

PR: https://git.openjdk.java.net/valhalla/pull/660

From vromero at openjdk.java.net  Fri Feb 25 18:19:42 2022
From: vromero at openjdk.java.net (Vicente Romero)
Date: Fri, 25 Feb 2022 18:19:42 GMT
Subject: Integrated: some test refactorings
In-Reply-To: <wVR7mK39PxqsVRdeqBwIhqs1nwrus4aDIpDTD9d2V48=.fee5a766-bd7d-465c-a799-6586101ad698@github.com>
References: <wVR7mK39PxqsVRdeqBwIhqs1nwrus4aDIpDTD9d2V48=.fee5a766-bd7d-465c-a799-6586101ad698@github.com>
Message-ID: <OVQOJv3hKBnAmBpYgKQ6rwkfadXA6g2tzpS61sGaHtQ=.84c4fa4f-6a82-48d1-a693-e15d0c047c03@github.com>

On Fri, 25 Feb 2022 18:12:34 GMT, Vicente Romero <vromero at openjdk.org> wrote:

> Refactoring tests at UniversalTvarsCompilationTests

This pull request has now been integrated.

Changeset: 02a49734
Author:    Vicente Romero <vromero at openjdk.org>
URL:       https://git.openjdk.java.net/valhalla/commit/02a49734bac4633e82ac32c34e5ec204349bcd59
Stats:     69 lines in 1 file changed: 8 ins; 39 del; 22 mod

some test refactorings

-------------

PR: https://git.openjdk.java.net/valhalla/pull/660