Hit NOTREACHED at

void ChildProcessTaskProvider::DeleteTask(base::ProcessHandle handle) {
  auto itr = tasks_by_handle_.find(handle);

  // The following case should never happen since we start observing
  // |BrowserChildProcessObserver| only after we collect all pre-existing child
  // processes and are notified (on the UI thread) that the collection is
  // completed at |ChildProcessDataCollected()|.
  if (itr == tasks_by_handle_.end()) {
    NOTREACHED();
    return;
  }


Flaked at r392913, r392926, r392926, r392941

[4114:4114:0511/073320:FATAL:child_process_task_provider.cc(147)] Check failed: false.
#0 0x000001a4b6ae base::debug::StackTrace::StackTrace()
#1 0x000001a5fcba logging::LogMessage::~LogMessage()
#2 0x0000018ff112 task_management::ChildProcessTaskProvider::DeleteTask()
#3 0x0000018fefd5 task_management::ChildProcessTaskProvider::BrowserChildProcessHostDisconnected()
#4 0x000002e24974 content::(anonymous namespace)::NotifyProcessHostDisconnected()
#5 0x000001aca509 base::debug::TaskAnnotator::RunTask()

posted from

BrowserChildProcessHostImpl::~BrowserChildProcessHostImpl() {
  g_child_process_list.Get().remove(this);

  if (notify_child_disconnected_) {
    BrowserThread::PostTask(BrowserThread::UI, FROM_HERE,
                            base::Bind(&NotifyProcessHostDisconnected, data_));
  }
}


This was added in r384454

commit 012289f
Author: amistry <amistry@chromium.org>
Date:   Fri Apr 1 11:56:37 2016
    Notify observers if a running browser child process is shut down by deleting the BrowserChildProcessHostImpl.
    BUG= 598775 
    Review URL: https://codereview.chromium.org/1843663003
    Cr-Commit-Position: refs/heads/master@{#384454}


I ran a git log on src/content looking for suspects - maybe things reverted shortly after r392941 but nothing stuck out. Maybe r392845 reverted in r392960?

miu's MediaStream audio: Refactor 3 separate "glue" implementations into one.

Not disabling any test for now.

This is almost certainly me. What I think is happening is that the utility process is started, and shut down before BrowserChildProcessHostImpl::OnChannelConnected() is run. This is not an error. The Safe JSON parser now uses Mojo and has no need to wait for Chrome-IPC to be initialised. So the task manager never gets the BrowserChildProcessLaunchedAndConnected event, but does get the BrowserChildProcessHostDisconnected event.

This only happens for very short lived processes. Since the task manager is the only thing that relies on BrowserChildProcessLaunchedAndConnected, the simplest solution is to delete that NOTREACHED() since it's not necessarily true that a process will be "connected" before it is destroyed. Maybe the observer semantics need to change to no longer indicate Chrome-IPC connection, since going forward, it will not be true (as we move to Mojo).

 Issue 613778  has been merged into this issue.

This is causing crashes on the debug Chrome OS Linux build very often. Could you please fix the BrowserChildProcessHostImpl somehow? I don't think we should remove that NOTREACHED(). It is there to expose problems exactly like this one.

 Issue 614254  has been merged into this issue.

I've started work on this.

Detected 3 new flakes for test/step "SafeJsonParserTest.Parse". To see the actual flakes, please visit https://chromium-try-flakes.appspot.com/all_flake_occurrences?key=ahVzfmNocm9taXVtLXRyeS1mbGFrZXNyIwsSBUZsYWtlIhhTYWZlSnNvblBhcnNlclRlc3QuUGFyc2UM. This message was posted automatically by the chromium-try-flakes app. Since flakiness is ongoing, the issue was moved back into Sheriff Bug Queue (unless already there).

Detected 6 new flakes for test/step "SafeJsonParserTest.Parse". To see the actual flakes, please visit https://chromium-try-flakes.appspot.com/all_flake_occurrences?key=ahVzfmNocm9taXVtLXRyeS1mbGFrZXNyIwsSBUZsYWtlIhhTYWZlSnNvblBhcnNlclRlc3QuUGFyc2UM. This message was posted automatically by the chromium-try-flakes app. Since flakiness is ongoing, the issue was moved back into Sheriff Bug Queue (unless already there).

Detected 5 new flakes for test/step "SafeJsonParserTest.Parse". To see the actual flakes, please visit https://chromium-try-flakes.appspot.com/all_flake_occurrences?key=ahVzfmNocm9taXVtLXRyeS1mbGFrZXNyIwsSBUZsYWtlIhhTYWZlSnNvblBhcnNlclRlc3QuUGFyc2UM. This message was posted automatically by the chromium-try-flakes app. Since flakiness is ongoing, the issue was moved back into Sheriff Bug Queue (unless already there).

Detected 3 new flakes for test/step "SafeJsonParserTest.Parse". To see the actual flakes, please visit https://chromium-try-flakes.appspot.com/all_flake_occurrences?key=ahVzfmNocm9taXVtLXRyeS1mbGFrZXNyIwsSBUZsYWtlIhhTYWZlSnNvblBhcnNlclRlc3QuUGFyc2UM. This message was posted automatically by the chromium-try-flakes app.

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/fc57dbf3f6a3526468babde9d77f7b56d3f2a68a

commit fc57dbf3f6a3526468babde9d77f7b56d3f2a68a
Author: amistry <amistry@chromium.org>
Date: Tue May 31 01:25:31 2016

Disable safe json browser test.

Test is flaky due to races with process startup/shutdown notifications.

BUG=611067
TBR=sky@chromium.org

Review-Url: https://codereview.chromium.org/2020123002
Cr-Commit-Position: refs/heads/master@{#396765}

[modify] https://crrev.com/fc57dbf3f6a3526468babde9d77f7b56d3f2a68a/chrome/browser/safe_json_parser_browsertest.cc

Removing from Sheriff queue since the test is disabled and the bug has an owner.

Moving this nonessential bug to the next milestone.

For more details visit https://www.chromium.org/issue-tracking/autotriage - Your friendly Sheriffbot

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/c7a521f9d0b1f589b9dafcfdac381d10be9a9258

commit c7a521f9d0b1f589b9dafcfdac381d10be9a9258
Author: amistry <amistry@chromium.org>
Date: Thu Jun 02 05:45:35 2016

Temorarily comment out NOTREACHED in task manager.

This condition is reached because of how Mojo interacts with child
process setup. While this is being fixed, comment out the check to avoid
test flakes on the bots.

BUG=611067

Review-Url: https://codereview.chromium.org/2025233002
Cr-Commit-Position: refs/heads/master@{#397296}

[modify] https://crrev.com/c7a521f9d0b1f589b9dafcfdac381d10be9a9258/chrome/browser/task_management/providers/child_process_task_provider.cc

This issue is Pri-1 but has already been moved once. Lowering the priority and moving to the next milestone.

For more details visit https://www.chromium.org/issue-tracking/autotriage - Your friendly Sheriffbot