[debugger][exception] Debugger breaks on handled exceptions #17106

thaystg · 2019-09-27T20:19:53Z

If there is a perform_wait_callback in the stack there will be another catch generated by the method called in the owner thread, so we don't need to throw as unhandled exception, we can continue and find the next catch.

Fixes #17083

…r catch generated by the owner thread, so we don't need to throw, we can continue and find the next catch. Fixes mono#17083

thaystg · 2019-09-28T12:41:48Z

@monojenkins build failed

thaystg · 2019-09-28T16:23:14Z

@monojenkins build failed

thaystg · 2019-09-28T21:41:16Z

@monojenkins build failed

marek-safar · 2019-09-30T08:11:37Z

@monojenkins build failed

marek-safar · 2019-09-30T08:11:45Z

@monojenkins backport 2019-08

vargaz · 2019-09-30T08:41:34Z

Is perform_wait_callback a special method so it needs to be special cased in the runtime ?

lambdageek · 2019-09-30T13:54:28Z

@vargaz PerformWaitCallback is a little bit special to the runtime. The threadpool code sets save_lmf to TRUE because the worker loop wants to mess with the exception handling.

mono/mono/metadata/threadpool.c

Lines 326 to 330 in 7f37f76

    
           	/* 
        
           	 * This is needed so there is always an lmf frame in the runtime invoke call below, 
        
           	 * so ThreadAbortExceptions are caught even if the thread is in native code. 
        
           	 */ 
        
           	mono_defaults.threadpool_perform_wait_callback_method->save_lmf = TRUE;

vargaz · 2019-10-01T08:53:14Z

I guess this is ok then but it still looks kinda weird.

knocte · 2019-10-01T11:37:20Z

I don't want to delay this fix to be committed (in fact I was the bug reporter), but hardcoding threadpool_perform_wait_callback_method in this bug fix would mean that the bug would still be present when the thread is not a threadpool thread?

thaystg · 2019-10-01T14:03:16Z

It's a side effect of #15234. We should backport to 2019-06 and 2019-02?

lambdageek · 2019-10-01T15:39:32Z

@monojenkins backport to 2019-06

monojenkins · 2019-10-01T15:40:36Z

@lambdageek backporting to 2019-06 failed, the patch results in conflicts:

Applying: If there is a perform_wait_callback in the stack there will be another catch generated by the owner thread, so we don't need to throw, we can continue and find the next catch. Fixes #17083
Using index info to reconstruct a base tree...
M	mcs/class/Mono.Debugger.Soft/Test/dtest-app.cs
M	mcs/class/Mono.Debugger.Soft/Test/dtest.cs
M	mono/mini/mini-exceptions.c
Falling back to patching base and 3-way merge...
Auto-merging mono/mini/mini-exceptions.c
CONFLICT (content): Merge conflict in mono/mini/mini-exceptions.c
Auto-merging mcs/class/Mono.Debugger.Soft/Test/dtest.cs
Auto-merging mcs/class/Mono.Debugger.Soft/Test/dtest-app.cs
error: Failed to merge in the changes.
Patch failed at 0001 If there is a perform_wait_callback in the stack there will be another catch generated by the owner thread, so we don't need to throw, we can continue and find the next catch. Fixes #17083

Please backport manually!

marek-safar · 2019-10-01T15:59:58Z

@monojenkins build failed

lambdageek · 2019-10-02T12:08:18Z

@monojenkins build failed

akoeplinger · 2019-10-02T17:26:30Z

@monojenkins build

akoeplinger · 2019-10-02T17:27:22Z

@monojenkins backport 2019-08

thaystg · 2019-10-02T18:13:08Z

@monojenkins build failed

thaystg · 2019-10-02T18:46:05Z

@monojenkins build failed

thaystg · 2019-10-02T19:55:06Z

@monojenkins build failed

thaystg · 2019-10-02T21:49:52Z

@monojenkins build failed

lambdageek · 2019-10-04T20:49:53Z

Relevant failures: x64 and arm64 FullAOT+Interpreter lanes consistently failing the abort-cctor test. Also on the 2019-08 backport.

…oline.

thaystg · 2019-10-07T02:28:18Z

@monojenkins build failed

thaystg · 2019-10-07T12:22:40Z

@monojenkins backport 2019-08

thaystg · 2019-10-07T12:23:09Z

@monojenkins backport 2019-06

monojenkins · 2019-10-07T12:24:09Z

@thaystg backporting to 2019-06 failed, the patch results in conflicts:

Applying: If there is a perform_wait_callback in the stack there will be another catch generated by the owner thread, so we don't need to throw, we can continue and find the next catch. Fixes #17083
Using index info to reconstruct a base tree...
M	mcs/class/Mono.Debugger.Soft/Test/dtest-app.cs
M	mcs/class/Mono.Debugger.Soft/Test/dtest.cs
M	mono/mini/mini-exceptions.c
Falling back to patching base and 3-way merge...
Auto-merging mono/mini/mini-exceptions.c
CONFLICT (content): Merge conflict in mono/mini/mini-exceptions.c
Auto-merging mcs/class/Mono.Debugger.Soft/Test/dtest.cs
Auto-merging mcs/class/Mono.Debugger.Soft/Test/dtest-app.cs
error: Failed to merge in the changes.
Patch failed at 0001 If there is a perform_wait_callback in the stack there will be another catch generated by the owner thread, so we don't need to throw, we can continue and find the next catch. Fixes #17083

Please backport manually!

…r catch generated by the method called in the owner thread, so we don't need to throw as unhandled exception, we can continue and find the next catch. Fixes mono#17083 Backport of mono#17106.

thaystg · 2019-10-07T12:36:05Z

@monojenkins backport 2019-02

monojenkins · 2019-10-07T12:51:08Z

@thaystg backporting to 2019-02 failed, the patch results in conflicts:

Applying: If there is a perform_wait_callback in the stack there will be another catch generated by the owner thread, so we don't need to throw, we can continue and find the next catch. Fixes #17083
Using index info to reconstruct a base tree...
M	mcs/class/Mono.Debugger.Soft/Test/dtest-app.cs
M	mcs/class/Mono.Debugger.Soft/Test/dtest.cs
M	mono/mini/mini-exceptions.c
Falling back to patching base and 3-way merge...
Auto-merging mono/mini/mini-exceptions.c
CONFLICT (content): Merge conflict in mono/mini/mini-exceptions.c
Auto-merging mcs/class/Mono.Debugger.Soft/Test/dtest.cs
Auto-merging mcs/class/Mono.Debugger.Soft/Test/dtest-app.cs
error: Failed to merge in the changes.
Patch failed at 0001 If there is a perform_wait_callback in the stack there will be another catch generated by the owner thread, so we don't need to throw, we can continue and find the next catch. Fixes #17083

Please backport manually!

…r catch generated by the method called in the owner thread, so we don't need to throw as unhandled exception, we can continue and find the next catch. Fixes mono#17083 Backport of mono#17106.

…r catch generated by the method called in the owner thread, so we don't need to throw as unhandled exception, we can continue and find the next catch. Fixes #17083 Backport of #17106.

…o#17106) * If there is a perform_wait_callback in the stack there will be another catch generated by the owner thread, so we don't need to throw, we can continue and find the next catch. Fixes mono/mono#17083 * Reverting unit test changed on commit 405d521. * Fixing assert when calling mono_jit_info_get_method if it was a trampoline. Commit migrated from mono/mono@18fac0a

If there is a perform_wait_callback in the stack there will be anothe…

342e4c4

…r catch generated by the owner thread, so we don't need to throw, we can continue and find the next catch. Fixes mono#17083

thaystg requested review from lewurm and vargaz as code owners September 27, 2019 20:19

Reverting unit test changed on commit 405d521.

04124aa

vargaz approved these changes Oct 1, 2019

View reviewed changes

monojenkins mentioned this pull request Oct 2, 2019

[2019-08] [debugger][exception] Debugger breaks on handled exceptions #17155

Closed

Fixing assert when calling mono_jit_info_get_method if it was a tramp…

b12bfb1

…oline.

thaystg requested a review from BrzVlad as a code owner October 7, 2019 01:20

thaystg merged commit 18fac0a into mono:master Oct 7, 2019

monojenkins mentioned this pull request Oct 7, 2019

[2019-08] [debugger][exception] Debugger breaks on handled exceptions #17202

Merged

thaystg mentioned this pull request Oct 7, 2019

[2019-06] [debugger][exception] Debugger breaks on handled exceptions #17203

Merged

monojenkins mentioned this pull request Oct 7, 2019

[2019-02] [2019-06] [debugger][exception] Debugger breaks on handled exceptions #17205

Closed

[debugger][exception] Debugger breaks on handled exceptions #17106

[debugger][exception] Debugger breaks on handled exceptions #17106

Uh oh!

Conversation

thaystg commented Sep 27, 2019 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

thaystg commented Sep 28, 2019

Uh oh!

thaystg commented Sep 28, 2019

Uh oh!

thaystg commented Sep 28, 2019

Uh oh!

marek-safar commented Sep 30, 2019

Uh oh!

marek-safar commented Sep 30, 2019

Uh oh!

vargaz commented Sep 30, 2019

Uh oh!

lambdageek commented Sep 30, 2019

Uh oh!

vargaz commented Oct 1, 2019

Uh oh!

knocte commented Oct 1, 2019

Uh oh!

thaystg commented Oct 1, 2019 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

lambdageek commented Oct 1, 2019

Uh oh!

monojenkins commented Oct 1, 2019

Uh oh!

marek-safar commented Oct 1, 2019

Uh oh!

lambdageek commented Oct 2, 2019

Uh oh!

akoeplinger commented Oct 2, 2019

Uh oh!

akoeplinger commented Oct 2, 2019

Uh oh!

thaystg commented Oct 2, 2019

Uh oh!

thaystg commented Oct 2, 2019

Uh oh!

thaystg commented Oct 2, 2019

Uh oh!

thaystg commented Oct 2, 2019

Uh oh!

lambdageek commented Oct 4, 2019

Uh oh!

thaystg commented Oct 7, 2019

Uh oh!

thaystg commented Oct 7, 2019

Uh oh!

thaystg commented Oct 7, 2019

Uh oh!

monojenkins commented Oct 7, 2019

Uh oh!

thaystg commented Oct 7, 2019

Uh oh!

monojenkins commented Oct 7, 2019

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

7 participants

thaystg commented Sep 27, 2019 •

edited

Loading

thaystg commented Oct 1, 2019 •

edited

Loading