8238812: assert(false) failed: bad AD file #2758

r-v-raghav · 2021-02-26T20:13:35Z

Found that the earlier PR bug analysis and fix details done is wrong.
Request help to review this revised fix proposal.

Now understood need to fix two types of issues related to the bug-

(i). Crash with only CmpI nodes case.
Sample test case -

public static void test2() {
  for (int i = 5; i > -20; i -= 5) {
    switch (i) {
      case 0:
      case 10:
      case 20:
      case 30:
      case 40:
      case 50:
      case 60:
      case 100:
        res = 42;
        break;
    }
  }
}

// This generates the following subgraph:

/*
// i: -20..0
if (i != 0) {
  // i: -20..-1
  if (i < 0) {   <- /*This is always true but not folded by C2*/
    // Fall through
  } else {
    ...
    CastII(i-1, 0..4) <- /*Replaced by TOP because i-1 range is -21..-1 but still considered reachable by C2 although it is dead code*/
    ...
  }
} else {
  StoreI   <- Due to this additional store on, IfNode::has_shared_region returns false and the fold compares optimization does not kick in
}
*/

Sample IR extracts -

 77  Phi  ===  399  25  389  [[ 658  664  389  80  662  660  659 ]]
 25  ConI  ===  0  [[ 80  77 ]]  #int:0
 80  CmpI  === _  77  25  [[ 81  89 ]]
 81  Bool  === _  80  [[ 82 ]] [ne]
 82  If  ===  399  81  [[ 85  87 ]]
 85  IfFalse  ===  82  [[ 102  336 ]]
 87  IfTrue  ===  82  [[ 90 ]] #1
 89  Bool  === _  80  [[ 90 ]] [lt]
 90  If  ===  87  89  [[ 91  92 ]]
 91  IfTrue  ===  90  [[ 102 ]] #1
 92  IfFalse  ===  90  [[ 156 ]] #0
 336  StoreI  ===  85  359  335  332  [[ 337 ]]
 156  Jump  ===  92  1  [[ 157  160  .......

Fix proposal here includes :

Extract the existing non-CmpU related optimization out of old fold_compares_helper(),
form new method
(new fold_dominated_if() added and renamed old fold_compares_helper as fold_to_unsigned()).
Call it properly from fold_compares() and related code adjustments.

(ii). fold_compares() not handling CmpU nodes cases.

Sample basic test case -

public static void test3() {
  int local = 0;
  for (int i = 5; i > -20; i -= 5) {
    switch (i) {
    case 0:
    case 10:
    case 20:
    case 30:
    case 40:
    case 50:
    case 100:
      local = 42;
      break;
    }
  }
}

// This generates the following subgraph:
/*
// i: -20..0
if (i != 0) {
  // i: -20..-1
  if (i u< 101) {    <- /*This is always false but not folded by C2 because CmpU is not handled*/
    CastII(i-1, 0..49)  <- /*Replaced by TOP because i-1 range is -21..-1 but still considered reachable by C2 although it is dead code*/
  } else {
    ...
  }
} else {
  ...
}
*/

Sample IR extracts -

 77  Phi  ===  363  20  353  [[ 573  577  572  80  575  353  345  574 ]]
 20  ConI  ===  0  [[ 80  77 ]]  #int:0
 80  CmpI  === _  77  20  [[ 81 ]]
 81  Bool  === _  80  [[ 82 ]] [ne]
 82  If  ===  363  81  [[ 85  87 ]]
 85  IfFalse  ===  82  [[ 102 ]] #0
 87  IfTrue  ===  82  [[ 337 ]] #1
 95  ConI  ===  0  [[ 345 ]]  #int:101
 345  CmpU  === _  77  95  [[ 346 ]]
 346  Bool  === _  345  [[ 337 ]] [lt]
 337  If  ===  87  346  [[ 338  339 ]]
 338  IfFalse  ===  337  [[ 102 ]] #0
 339  IfTrue  ===  337  [[ 135 ]] #1
 135  Jump  ===  339  1  [[ 136  139  142  .....

Proposed fix includes:
Enabling Op_CmpU support in -

cmpi_folds(),
filtered_int_type() - returns possible restrictive type for input value based on condition control flow for the if (supported CmpU for various condition control types).
fold_dominated_if() - checks if dominating if determines the result of current if and above example test case is a basic variant possibility where both CmpNodes compare a common value directly. (Node 77 Phi in above sample). But as mentioned in the comments of new helper function get_base_comparing_value(), there are other possibilities where there might be an AddINode with a constant increment present in between the CmpNodes and the common value compared.
(Sample IR extract with AddI in between CmpNodes : )

 76  Phi  ===  364  79  354  [[ 576  580  575  80  578  354  344  577 ]]  #int:0..20:www
 79  ConI  ===  0  [[ 80  76 ]]  #int:20
 80  CmpI  === _  76  79  [[ 81 ]]
 81  Bool  === _  80  [[ 82 ]] [ne] 
 82  If  ===  364  81  [[ 85  87 ]]
 85  IfFalse  ===  82  [[ 102 ]] #0
 87  IfTrue  ===  82  [[ 338 ]] #1
 343  ConI  ===  0  [[ 344 ]]  #int:-20
 344  AddI  === _  76  343  [[ 348 ]] 
 347  ConI  ===  0  [[ 348 ]]  #int:281
 348  CmpU  === _  344  347  [[ 349 ]] 
 349  Bool  === _  348  [[ 338 ]] [lt]
 338  If  ===  87  349  [[ 339  340 ]]
 339  IfFalse  ===  338  [[ 102 ]] #0 
 340  IfTrue  ===  338  [[ 136 ]] #1
 136  Jump  ===  340  1  [[ 137  140  143  146 .........

All these variants are handled in get_base_comparing_value() and related test cases added.

and related changes in fold_compares().

Found no issues with this proposed fix for various unit tests, reported java_fuzzer case, tier1-7 and pre-integeration tests

Progress

Change must not contain extraneous whitespace
Commit message must refer to an issue
Change must be properly reviewed

Issue

JDK-8238812: assert(false) failed: bad AD file

Download

$ git fetch https://git.openjdk.java.net/jdk pull/2758/head:pull/2758
$ git checkout pull/2758

Progress

Change must not contain extraneous whitespace
Commit message must refer to an issue
Change must be properly reviewed

Issue

JDK-8238812: assert(false) failed: bad AD file

Reviewers

Tobias Hartmann (@TobiHartmann - Reviewer)
Christian Hagedorn (@chhagedorn - Reviewer)
Roland Westrelin (@rwestrel - Reviewer)

Download

$ git fetch https://git.openjdk.java.net/jdk pull/2758/head:pull/2758
$ git checkout pull/2758

bridgekeeper · 2021-02-26T20:14:28Z

👋 Welcome back rraghavan! A progress list of the required criteria for merging this PR into master will be added to the body of your pull request. There are additional pull request commands available for use with this pull request.

openjdk · 2021-02-26T20:15:44Z

@r-v-raghav The following label will be automatically applied to this pull request:

hotspot-compiler

When this pull request is ready to be reviewed, an "RFR" email will be sent to the corresponding mailing list. If you would like to change these labels, use the /label pull request command.

TobiHartmann · 2021-03-01T06:50:15Z

src/hotspot/share/opto/ifnode.cpp

              hi = TypeInt::INT->_hi;
              break;
            default:
+              assert(false, "should not reach here");


Use ShouldNotReachHere(); instead.

TobiHartmann · 2021-03-01T06:54:21Z

src/hotspot/share/opto/ifnode.cpp

+// in-between the CmpNode(s) and the common value we compare.
+// Check for the following cases. Then return common value compared
+// if present and also save the constant values to be adjusted or
+// subtracted due to the possible AddINode in-between.


I would replace adjusted or subtracted and say something like ... and also save the constant value that is added to infer the type of the common value we compare.

TobiHartmann · 2021-03-01T06:54:51Z

src/hotspot/share/opto/ifnode.cpp

+
+  Node* res_val = NULL;
+  this_adj_val = 0;
+  dom_adj_val = 0;


The caller already initialized these to 0, right?

TobiHartmann · 2021-03-01T06:57:56Z

src/hotspot/share/opto/ifnode.cpp

+Node* IfNode::get_base_comparing_value (Node* dom_if, PhaseIterGVN* igvn, jint& this_adj_val, jint& dom_adj_val) {
+  Node* this_cmp = in(1)->in(1);
+  Node* dom_cmp = dom_if->in(1)->in(1);
+  assert(this_cmp->is_Cmp() != false && dom_cmp->is_Cmp() != false, "compare expected");


No need to compare against false. This should be assert(this_cmp->is_Cmp() && dom_cmp->is_Cmp(), ... but instead you could also simply replace declarations above by Node* this_cmp = in(1)->in(1)->as_Cmp() which would include the assert.

TobiHartmann · 2021-03-01T06:58:25Z

src/hotspot/share/opto/ifnode.cpp

+  Node* dom_val = dom_cmp->in(1);
+  Node* this_val = this_cmp->in(1);
+  if (this_val == dom_val) {
+    res_val = this_val;


Add comments describing that this is Variant 1, same for other if branches below.

TobiHartmann · 2021-03-01T07:03:33Z

src/hotspot/share/opto/ifnode.cpp

+    res_val = this_val;
+  } else if (this_val->is_Add() && this_val->in(1) && this_val->in(1) == dom_val) {
+    const TypeInt* val_t = igvn->type(this_val->in(2))->isa_int();
+    if (val_t && val_t->is_con()) {


val_t implicitly casts a pointer to a boolean, should be replaced by val_t != NULL. Same below.

TobiHartmann · 2021-03-01T07:05:17Z

src/hotspot/share/opto/ifnode.cpp

+  jint dom_adj_val = 0;
+  Node* n = NULL;
+
+  n = get_base_comparing_value(dom_if, igvn, this_adj_val, dom_adj_val);


No need to initialize n above. Should be Node* n = get_base_comparing_value(...

Actually, you are not even using n and can therefore just return a boolean from get_base_comparing_value and simplify that method accordingly.

TobiHartmann · 2021-03-01T07:07:14Z

src/hotspot/share/opto/ifnode.cpp

+
+// Check if dominating if determines the result of this if
+bool IfNode::fold_dominated_if(ProjNode* proj, PhaseIterGVN* igvn) {
+  Node* this_cmp = in(1)->in(1);


We only ever use this_cmp->in(1) so you can replace this by Node* this_val = in(1)->in(1)->in(1). Same for dom_cmp below.

TobiHartmann · 2021-03-01T07:09:18Z

src/hotspot/share/opto/ifnode.cpp

+      failtype = dom_cmp->in(1)->as_Add()->add_ring(failtype, TypeInt::make(-dom_adj_val))->is_int();
+    }
+    for (int i = 0; i < 2; ++i) {
+      const TypeInt* type2 = filtered_int_type(igvn, this_cmp->in(1), proj_out(i));


Why not just name this type?

TobiHartmann · 2021-03-01T07:09:59Z

src/hotspot/share/opto/ifnode.cpp

+          assert(this_cmp->in(1)->is_Add() != false, "sanity");
+          type2 = this_cmp->in(1)->as_Add()->add_ring(type2, TypeInt::make(-this_adj_val))->is_int();
+        }
+        const TypeInt* res_type = failtype->join(type2)->is_int();


No need to declare res_type, just use type.

mlbridge · 2021-03-02T08:20:15Z

Webrevs

chhagedorn · 2021-03-02T09:26:25Z

src/hotspot/share/opto/ifnode.cpp

+  Node* this_cmp = in(1)->in(1)->as_Cmp();
+  Node* dom_cmp = dom_if->in(1)->in(1)->as_Cmp();


Could be directly turned into an assertion with is_Cmp() instead of using as_Cmp().

chhagedorn · 2021-03-02T09:31:18Z

src/hotspot/share/opto/ifnode.cpp

+  if (this_val == dom_val) {
+    // Variant 1
+    return true;
+  } else if (this_val->is_Add() && this_val->in(1) != NULL && this_val->in(1) == dom_val) {


As this_val and dom_val are both non-null: this_val->in(1) != NULL is always true when this_val->in(1) == dom_val holds and thus this_val->in(1) != NULL can be removed. This can also be applied in Variant 3.

chhagedorn · 2021-03-02T09:42:27Z

src/hotspot/share/opto/ifnode.cpp

+      dom_adj_val = val_t->get_con();
+      return true;
+    }
+  } else if (this_val->is_Add() && dom_val->is_Add() && this_val->in(1) != NULL && dom_val->in(1) != NULL && this_val->in(1) == dom_val->in(1)) {


Here I think you can just remove one of the non-null checks (in case both are NULL).

chhagedorn · 2021-03-02T09:47:33Z

src/hotspot/share/opto/ifnode.cpp

@@ -837,8 +857,94 @@ bool IfNode::has_only_uncommon_traps(ProjNode* proj, ProjNode*& success, ProjNod
  return false;
 }

+// There might be an AddINode (marked with *) with a constant increment


Do we only need to do this for AddNodes? What about SubNodes for example?

Never found a SubNode in this case. Always AddI node with integer constant as second input (this constant
can be positive or negative values)

That's because we always convert SubINodes with constant second operand to AddINodes:

jdk/src/hotspot/share/opto/subnode.cpp

Line 181 in 3d3eb5c

// Convert "x-c0" into "x+ -c0".

Okay I see, thanks for the explanation.

chhagedorn · 2021-03-02T10:34:48Z

src/hotspot/share/opto/ifnode.cpp

+              if (is_unsigned && lo >= 0) {
+                // val u< cmp2 only passes for val >= 0 if cmp2 >= 0
+                lo = 0;
+              } else {
+                lo = TypeInt::INT->_lo;
+              }


Had some problems following this logic. But I think I got it now. What do you think about reformulating this (and the related comments further down) into

if (is_unsigned && lo >= 0) { // cmp2 >= 0: val u<= cmp2 can only pass if val >= 0. Set val->_lo to 0. } else { // The lower bound of val cannot be improved. lo = TypeInt::INT->_lo; }

to make it more clear?

chhagedorn · 2021-03-02T10:43:35Z

test/hotspot/jtreg/compiler/c2/TestJumpTable.java

+
+    // Original (slightly simplified) fuzzer generated test
+    public static void test1() {
+        int i4, i5=99, i6, i9=89;


Minor: Add spaces

chhagedorn · 2021-03-02T11:11:09Z

src/hotspot/share/opto/ifnode.cpp

@@ -700,16 +720,17 @@ const TypeInt* IfNode::filtered_int_type(PhaseGVN* gvn, Node* val, Node* if_proj
 //

 // Is the comparison for this If suitable for folding?
-bool IfNode::cmpi_folds(PhaseIterGVN* igvn, bool fold_ne) {
+bool IfNode::cmpi_folds(PhaseIterGVN* igvn) {


We should probably rename it into cmp_folds now that it is also for unsigned and also update the description above. Shouldn't the description be at fold_compares() anyways? But maybe we want to clean this up in a follow-up RFE.

Yes, will rename cmpi_folds to cmp_folds.

Not changing the fold compares description location for now.
For now maybe it is okay location for description as that is starting of fold_compares implementation (all methods above fold_compares definition seems mainly its helper functions)
.e.g.: See the comment and declarations in [cfgnode.hpp]
........
private:
// Helper methods for fold_compares
bool cmp_folds(PhaseIterGVN* igvn);
bool is_ctrl_folds(Node* ctrl, PhaseIterGVN* igvn);
Or yes agree we can change this in a follow-up RFE.

Yeah, it's probably best to leave it as it is at the moment and clean it up later in an RFE.

r-v-raghav · 2021-03-03T11:27:41Z

Thanks to Tobias and Christian for all help review comments, suggestions.
Please note I could address all the comments suggestions above so far.

TobiHartmann

This looks good to me.

openjdk · 2021-03-03T14:15:46Z

@r-v-raghav This change now passes all automated pre-integration checks.

ℹ️ This project also has non-automated pre-integration requirements. Please see the file CONTRIBUTING.md for details.

After integration, the commit message for the final commit will be:

8238812: assert(false) failed: bad AD file

Reviewed-by: thartmann, chagedorn, roland

You can use pull request commands such as /summary, /contributor and /issue to adjust it as needed.

At the time when this comment was updated there had been 134 new commits pushed to the master branch:

b7f0b3f: 8252173: Use handles instead of jobjects in modules.cpp
a6e34b3: 8262829: Native crash in Win32PrintServiceLookup.getAllPrinterNames()
fbe40e8: 8252399: Update mapMulti documentation to use type test pattern instead of instanceof once JEP 375 exits preview
0f2402d: 8263190: Update java.io, java.math, and java.text to use instanceof pattern variable
4f0a12e: 8262323: do not special case JVMCI in tiered compilation policy
3022baa: 8263167: IGV: build fails with "taskdef AutoUpdate cannot be found"
0bc4562: 8263068: Rename safefetch.hpp to safefetch.inline.hpp
5bfc5fd: 8263051: Modernize the code in the java.awt.color package
5b9b170: 8262955: Unify os::fork_and_exec() across Posix platforms
39b1113: 8262161: Refactor manual I/O stream copying in java.desktop to use new convenience APIs
... and 124 more: https://git.openjdk.java.net/jdk/compare/240f2a1bb7c982f7bda15c62abe2313d87654ed5...master

As there are no conflicts, your changes will automatically be rebased on top of these commits when integrating. If you prefer to avoid this automatic rebasing, please check the documentation for the /integrate command for further details.

➡️ To integrate this PR with the above commit message to the master branch, type /integrate in a new comment.

chhagedorn

Looks good to me!

chhagedorn · 2021-03-04T09:02:47Z

src/hotspot/share/opto/ifnode.cpp

@@ -837,8 +857,94 @@ bool IfNode::has_only_uncommon_traps(ProjNode* proj, ProjNode*& success, ProjNod
  return false;
 }

+// There might be an AddINode (marked with *) with a constant increment


Okay I see, thanks for the explanation.

rwestrel · 2021-03-05T14:24:07Z

src/hotspot/share/opto/ifnode.cpp

+              } else {
+                // The lower bound of val cannot be improved.
+                lo = TypeInt::INT->_lo;
+              }
              if (hi != min_jint) {
                hi = hi - 1;
              }
              break;


I think there's a bug here (case BoolTest::lt, unsigned comparison) . if cmp2 is [-1, 1] for instance, then lo and hi are unchanged. If the type of val is TypeInt::INT then the resulting type once join is applied is the interval [-1, 1]. But for instance 42 <u -1, 42 is in TypeInt::INT, -1 is in [-1, 1] but 42 is not in the resulting interval. I wonder if other cases in this method has similar issues. Can you fix that one and verify others?

rwestrel · 2021-03-05T14:35:12Z

This fixes the bug when the tests that need to be optimized out are next to one an another and the logic for folding comparisons can be used. It's hard to be certain that it's not possible for the 2 tests be separated by some other control flow. In that case the fix would fail.

A conservative fix would be to not use a narrow type in the CastII that tableswitch/lookupswitch switch introduces. It would be a simpler fix and a more robust one.

r-v-raghav · 2021-03-09T06:34:26Z

Thank you @rwestrel for the comments.
Understood your point.
Based on the review comments now planning to -

i. Do proposal by Roland and remove the narrow type from the CastII as fix for this bug task.
As of now found no issues with reported java_fuzzer cases / unit tests, mach5 tier 1-7 & pre-integration tests
for the following changes to avoid use any narrow type in the CastII introduced by lookupswitch.

[src/hotspot/share/opto/parse2.cpp]

   // Clean the 32-bit int into a real 64-bit offset.
   // Otherwise, the jint value 0 might turn into an offset of 0x0800000000.
-  const TypeInt* ikeytype = TypeInt::make(0, num_cases, Type::WidenMin);
   // Make I2L conversion control dependent to prevent it from
   // floating above the range check during loop optimizations.
-  key_val = C->conv_I2X_index(&_gvn, key_val, ikeytype, control());
+  // Do not use a narrow int type here to prevent the data path from dying
+  // while the control path is not removed. This can happen if the type of key_val
+  // is later known to be out of bounds of [0, num_cases] and therefore a narrow cast
+  // would be replaced by TOP while C2 is not able to fold the corresponding range checks.
+#ifdef _LP64
+  key_val = C->constrained_convI2L(&_gvn, key_val, TypeInt::INT, control());
+#endif

ii. Integrate the other proposed fix to improve fold_compares optimizations as a separate enhancement.
(after addressing the comments in filtered_int_type())

r-v-raghav · 2021-03-09T09:41:03Z

Please note as per the above comments now updated the fix. Request help with review. Thanks.

TobiHartmann

This looks good to me. Please file a follow-up enhancement to integrate the optimization, I think it's still valuable.

TobiHartmann

After quickly discussing this with @rwestrel, it turned out that you need to set CastIINode::_carry_dependency to prevent the CastIINode with TypeInt::INT from being eliminated and re-introducing (JDK-6675699).

r-v-raghav · 2021-03-09T14:45:40Z

Thanks for the comments @TobiHartmann , @rwestrel .

Please note now revised fix to add a _carry_dependeny argument to constrained_convI2L and set it to true for the CastII introduced by lookupswitch.
(found no issues with reported java_fuzzer cases / unit tests and
tier & pre-integration tests in progress)

Also created followup enhancement task - JDK-8263252.
Please let me know if missed anything.

rwestrel

Thanks for updating the change.

TobiHartmann

Looks good to me too.

chhagedorn

Looks good!

r-v-raghav · 2021-03-10T15:58:39Z

Thanks to @rwestrel @TobiHartmann @chhagedorn for review approvals.

(Now testing done - java_fuzzer cases / unit tests, mach5 tier 1-7 & pre-integration tests)

r-v-raghav · 2021-03-10T15:58:56Z

/integrate

openjdk · 2021-03-10T15:59:59Z

@r-v-raghav Since your change was applied there have been 149 commits pushed to the master branch:

b2a2ddf: 8262438: sun/security/ssl/SSLLogger/LoggingFormatConsistency.java failed with "SocketException: Socket is closed"
c8c0234: 8262471: Fix coding style in src/java.base/share/classes/java/lang/CharacterDataPrivateUse.java
4d21a45: 8262913: KlassFactory::create_from_stream should never return NULL
fab5676: 8247869: Change NONCOPYABLE to delete the operations
c0542ed: 6251901: BasicTextUI: installDefaults method are contrary to the documentation
fdd3941: 8263233: Update java.net and java.nio to use instanceof pattern variable
3fe8a46: 8263170: ComboBoxModel documentation refers to a nonexistent type
d8a9c3c: 8263002: Remove CDS MiscCode region
67ea3bd: 8263102: Expand documention of Method.isBridge
d0c1aec: 8263140: Japanese chars garble in console window in HSDB
... and 139 more: https://git.openjdk.java.net/jdk/compare/240f2a1bb7c982f7bda15c62abe2313d87654ed5...master

Your commit was automatically rebased without conflicts.

Pushed as commit 4b5be40.

💡 You may see a message that your pull request was closed with unmerged commits. This can be safely ignored.

8238812: assert(false) failed: bad AD file

8ee73a7

openjdk bot added the hotspot-compiler hotspot-compiler-dev@openjdk.org label Feb 26, 2021

8238812: assert(false) failed: bad AD file

eff906a

TobiHartmann suggested changes Mar 1, 2021

View reviewed changes

8238812: assert(false) failed: bad AD file

90ac1bd

r-v-raghav marked this pull request as ready for review March 2, 2021 08:15

openjdk bot added the rfr Pull request is ready for review label Mar 2, 2021

chhagedorn suggested changes Mar 2, 2021

View reviewed changes

8238812: assert(false) failed: bad AD file

5cef73a

TobiHartmann approved these changes Mar 3, 2021

View reviewed changes

openjdk bot added the ready Pull request is ready to be integrated label Mar 3, 2021

chhagedorn approved these changes Mar 4, 2021

View reviewed changes

rwestrel suggested changes Mar 5, 2021

View reviewed changes

8238812: assert(false) failed: bad AD file

b80e1b6

TobiHartmann approved these changes Mar 9, 2021

View reviewed changes

TobiHartmann suggested changes Mar 9, 2021

View reviewed changes

8238812: assert(false) failed: bad AD file

b027ab4

rwestrel approved these changes Mar 9, 2021

View reviewed changes

TobiHartmann approved these changes Mar 9, 2021

View reviewed changes

chhagedorn approved these changes Mar 9, 2021

View reviewed changes

openjdk bot closed this Mar 10, 2021

openjdk bot added integrated Pull request has been integrated and removed ready Pull request is ready to be integrated rfr Pull request is ready for review labels Mar 10, 2021

		Node* this_cmp = in(1)->in(1)->as_Cmp();
		Node* dom_cmp = dom_if->in(1)->in(1)->as_Cmp();

8238812: assert(false) failed: bad AD file #2758

8238812: assert(false) failed: bad AD file #2758

Conversation

r-v-raghav commented Feb 26, 2021 • edited by openjdk bot Loading

Progress

Issue

Download

Progress

Issue

Reviewers

Download

bridgekeeper bot commented Feb 26, 2021

openjdk bot commented Feb 26, 2021

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

mlbridge bot commented Mar 2, 2021 • edited Loading

Webrevs

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

r-v-raghav commented Mar 3, 2021 • edited Loading

TobiHartmann left a comment

Choose a reason for hiding this comment

openjdk bot commented Mar 3, 2021 • edited Loading

chhagedorn left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

rwestrel commented Mar 5, 2021

r-v-raghav commented Mar 9, 2021 • edited Loading

r-v-raghav commented Mar 9, 2021

TobiHartmann left a comment

Choose a reason for hiding this comment

TobiHartmann left a comment

Choose a reason for hiding this comment

r-v-raghav commented Mar 9, 2021

rwestrel left a comment

Choose a reason for hiding this comment

TobiHartmann left a comment

Choose a reason for hiding this comment

chhagedorn left a comment

Choose a reason for hiding this comment

r-v-raghav commented Mar 10, 2021

r-v-raghav commented Mar 10, 2021

openjdk bot commented Mar 10, 2021

r-v-raghav commented Feb 26, 2021 •

edited by openjdk bot

Loading

mlbridge bot commented Mar 2, 2021 •

edited

Loading

r-v-raghav commented Mar 3, 2021 •

edited

Loading

openjdk bot commented Mar 3, 2021 •

edited

Loading

r-v-raghav commented Mar 9, 2021 •

edited

Loading