New issue
Advanced search Search tips
Note: Color blocks (like or ) mean that a user may not be available. Tooltip shows the reason.

Issue 823488 link

Starred by 2 users

Issue metadata

Status: Fixed
Owner:
please use my google.com address
Closed: Mar 2018
Cc:
Components:
EstimatedDays: ----
NextAction: ----
OS: Chrome
Pri: 1
Type: Bug-Regression



Sign in to add a comment

ChromeOS chrome_user_log is full of "not implemented" error log lines

Project Member Reported by w...@chromium.org, Mar 19 2018

Issue description

Chrome Version: 66.0.3359.31
OS: ChromeOS Chromebook Plus

What steps will reproduce the problem?
(1) Sign-in to device and use it for a while.
(2) Open chrome://system
(3) Open the chrome_user_log

What is the expected result?

Expect that the only entries logged there are infrequent informational status logs, and infrequent/unusual error situations.

What happens instead?

Log is full of "not implemented" error entries, e.g:

[14003:14021:0319/145020.203751:ERROR:upload_data_presenter.cc(73)] Not implemented reached in virtual void extensions::RawDataPresenter::FeedNext(const net::UploadElementReader &)

[14003:14003:0319/131255.659208:ERROR:views_screen_locker.cc(114)] Not implemented reached in virtual void chromeos::ViewsScreenLocker::SetPasswordInputEnabled(bool)

[14003:14003:0319/131255.709547:ERROR:views_screen_locker.cc(131)] Not implemented reached in virtual void chromeos::ViewsScreenLocker::AnimateAuthenticationSuccess()

Aside from possibly indicating issues with this version of the build, this error logging eats up log space and makes it hard to see actual error logging.
 

Comment 1 by w...@chromium.org, Mar 19 2018

Cc: amistry@chromium.org
Components: Internals>Network>Proxy Internals>Plugins>Flash
Some other errors logged, that may indicate underlying issues that need resolving before stable:

[14003:14021:0319/150439.352936:ERROR:service_manager.cc(156)] The Service Manager prevented service "content_browser" from binding interface "content::mojom::ResourceUsageReporter" in target service "nacl_loader". You probably need to update one or more service manifests to ensure that "nacl_loader" exposes "content::mojom::ResourceUsageReporter" through a capability and that "content_browser" requires that capability from the "nacl_loader" service.

[14003:14003:0319/134447.510383:ERROR:pepper_flash_component_installer.cc(94)] Component flash registration failed


Comment 2 by w...@chromium.org, Mar 19 2018

Cc: -amistry@chromium.org ben@chromium.org

Comment 3 by mmenke@chromium.org, Mar 19 2018

Neither of those errors looks like they have anything to do with network proxies.  Did you add the Proxy label by mistake?

Comment 4 by w...@chromium.org, Mar 19 2018

Components: Internals>Services>ServiceManager
Re #3: No, but the ResourceUsageReporter stuff looks like it was added by amistry@ for the purpose of reporting JS memory usage in utility processes. Looks like ben@ moved that interface to content.mojom, so perhaps that is why it is now failing, so have CC'd him and added ServiceManager component.

Comment 5 by mmenke@chromium.org, Mar 19 2018

Still not seeing how any of that has anything to do with components/proxy, net/proxy, or any ChromeOS-specific proxy magic.

Comment 6 by w...@chromium.org, Mar 19 2018

Re #5: That tag was copied based on the bug in which the ResourceUsageReporting interface was added, on the assumption that the logged message mentioning it means it is in some way related to that.

I assume that the rationale for tagging that bug with Proxy was that proxy resolution in general involves PAC processing, for which we spawn a utility process to run the JS.

Feel free to remove the Proxy tag from this bug if you think it's unrelated!
Cc: rdevlin....@chromium.org
Owner: roc...@chromium.org
Status: Assigned (was: Untriaged)
Ken, looks like you're working in this area (and in fact some of the not implemented's were removed in revision 818bea4e4f6a919a93f3af3e2375f69099d2dfe2).  Is this already fixed (or in the works)?

Comment 8 by roc...@chromium.org, Mar 23 2018

Yep seems like a bug that is my fault. I'll take a look.

Comment 9 by roc...@chromium.org, Mar 26 2018

The ResourceUsageReporter error does just look like an incomplete interface rename. That should be its own bug, and I'll take care of that separately.

The not-implemented spam has likely only moved from [1] to [2]. I don't think any work I've done on network-service WebRequest could be relevant here though, since that log spam comes from the non-network-service code path and the logic hasn't changed for some time.

More likely this should simply not be a NOTIMPLEMENTED, but rather a DVLOG at best: it is reasonable for a URL request to have upload data attached which is neither of "file" or "bytes" type.

[1] https://cs.chromium.org/chromium/src/extensions/browser/api/web_request/upload_data_presenter.cc?rcl=5fcad4939c2de514758faa2d3735a4729af65101&l=73

[2] https://cs.chromium.org/chromium/src/extensions/browser/api/web_request/web_request_info.cc?rcl=760a5842b886b78538b37aad42965f277e564f4b&l=151

Project Member

Comment 10 by bugdroid1@chromium.org, Mar 27 2018

The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/92a05bc427badcfccd8f15ab10eb3b133d78e2ad

commit 92a05bc427badcfccd8f15ab10eb3b133d78e2ad
Author: Ken Rockot <rockot@chromium.org>
Date: Tue Mar 27 03:35:39 2018

WebRequest upload data: NOTIMPLEMENTED() -> DVLOG(1)

WebRequest API only supports exposing information about file and raw
byte upload data types. URL requests which include other types of upload
data should not result in log spam.

Bug:  823488 
Change-Id: I9ab9a680bf747dd0867809a6c3aed6a996b8d175
Reviewed-on: https://chromium-review.googlesource.com/980747
Reviewed-by: Devlin <rdevlin.cronin@chromium.org>
Commit-Queue: Ken Rockot <rockot@chromium.org>
Cr-Commit-Position: refs/heads/master@{#545980}
[modify] https://crrev.com/92a05bc427badcfccd8f15ab10eb3b133d78e2ad/extensions/browser/api/web_request/web_request_info.cc

Status: Fixed (was: Assigned)

Comment 12 by josa...@google.com, Mar 27 2018

Is this safe to merge to M-66?
Yes, safe to merge

Comment 14 by josa...@google.com, Mar 28 2018

Labels: Merge-Approved-66
ok, approved for merge to m66
Project Member

Comment 15 by bugdroid1@chromium.org, Mar 28 2018

Labels: -merge-approved-66 merge-merged-3359
The following revision refers to this bug:
  https://chromium.googlesource.com/chromium/src.git/+/97706a827617df134e8095f00d7f01cefa16bdfb

commit 97706a827617df134e8095f00d7f01cefa16bdfb
Author: Ken Rockot <rockot@chromium.org>
Date: Wed Mar 28 14:13:20 2018

Remove log spam from WebRequest API

This is backport of https://crrev.com/545980 since the original log
spam moved after M66 branch.

TBR=rdevlin.cronin@chromium.org

Bug:  823488 
Change-Id: I82384f055c2f0d671c576c972213ca8fc06ee279
Reviewed-on: https://chromium-review.googlesource.com/984032
Reviewed-by: Ken Rockot <rockot@chromium.org>
Cr-Commit-Position: refs/branch-heads/3359@{#483}
Cr-Branched-From: 66afc5e5d10127546cc4b98b9117aff588b5e66b-refs/heads/master@{#540276}
[modify] https://crrev.com/97706a827617df134e8095f00d7f01cefa16bdfb/extensions/browser/api/web_request/upload_data_presenter.cc

Sign in to add a comment