272,000 instances of "pipe error: 109" emitted from ipc/chromium/src/chrome/common/ipc_channel_win.cc during windows10-64 debug testing
Categories
(Core :: IPC, defect, P1)
Tracking
()
Tracking | Status | |
---|---|---|
firefox72 | --- | fixed |
People
(Reporter: erahm, Assigned: jld)
References
(Blocks 1 open bug)
Details
Attachments
(1 file)
272025 WARNING: pipe error: 109: file ipc/chromium/src/chrome/common/ipc_channel_win.cc, line 341
This warning [1] shows up in the following test suites:
16714 - test-windows10-64/debug-web-platform-tests-e10s-12 wpt12
14773 - test-windows10-64/debug-web-platform-tests-e10s-17 wpt17
13785 - test-windows10-64/debug-web-platform-tests-e10s-14 wpt14
13648 - test-windows10-64/debug-web-platform-tests-e10s-18 wpt18
13436 - test-windows10-64/debug-web-platform-tests-e10s-15 wpt15
13330 - test-windows10-64/debug-web-platform-tests-e10s-16 wpt16
13251 - test-windows10-64/debug-web-platform-tests-e10s-5 wpt5
13138 - test-windows10-64/debug-web-platform-tests-e10s-8 wpt8
12151 - test-windows10-64/debug-web-platform-tests-e10s-7 wpt7
11313 - test-windows10-64/debug-web-platform-tests-e10s-11 wpt11
11152 - test-windows10-64/debug-web-platform-tests-e10s-10 wpt10
10518 - test-windows10-64/debug-web-platform-tests-e10s-9 wpt9
10298 - test-windows10-64/debug-web-platform-tests-e10s-3 wpt3
9941 - test-windows10-64/debug-web-platform-tests-e10s-6 wpt6
8808 - test-windows10-64/debug-web-platform-tests-e10s-4 wpt4
8512 - test-windows10-64/debug-web-platform-tests-e10s-2 wpt2
8309 - test-windows10-64/debug-test-verify-e10s-1 TV1
7113 - test-windows10-64/debug-web-platform-tests-e10s-1 wpt1
6438 - test-windows10-64/debug-web-platform-tests-wdspec-e10s-2 Wd2
3894 - test-windows10-64/debug-web-platform-tests-wdspec-e10s-1 Wd1
2923 - test-windows10-64/debug-web-platform-tests-reftests-e10s-4 Wr4
2682 - test-windows10-64/debug-web-platform-tests-reftests-e10s-1 Wr1
2546 - test-windows10-64/debug-web-platform-tests-reftests-e10s-3 Wr3
2506 - test-windows10-64/debug-reftest-e10s-3 R3
2491 - test-windows10-64/debug-test-verify-wpt-e10s-1 TVw1
2466 - test-windows10-64/debug-test-verify-wpt-e10s-2 TVw2
2449 - test-windows10-64/debug-reftest-e10s-2 R2
2448 - test-windows10-64/debug-mochitest-browser-chrome-e10s-5 bc5
2177 - test-windows10-64/debug-mochitest-browser-chrome-e10s-6 bc6
2172 - test-windows10-64/debug-web-platform-tests-reftests-e10s-2 Wr2
2164 - test-windows10-64/debug-web-platform-tests-reftests-e10s-5 Wr5
2101 - test-windows10-64/debug-mochitest-browser-chrome-e10s-4 bc4
2015 - test-windows10-64/debug-mochitest-browser-chrome-e10s-3 bc3
1986 - test-windows10-64/debug-test-verify-wpt-e10s-3 TVw3
1974 - test-windows10-64/debug-marionette-e10s Mn
1921 - test-windows10-64/debug-mochitest-browser-chrome-e10s-1 bc1
1538 - test-windows10-64/debug-mochitest-e10s-5 5
1363 - test-windows10-64/debug-mochitest-e10s-3 3
1235 - test-windows10-64/debug-mochitest-e10s-4 4
1201 - test-windows10-64/debug-test-verify-gpu-e10s TVg
1168 - test-windows10-64/debug-mochitest-e10s-2 2
1062 - test-windows10-64/debug-mochitest-e10s-1 1
1023 - test-windows10-64/debug-firefox-ui-functional-local-e10s en-US
735 - test-windows10-64/debug-mochitest-devtools-chrome-e10s-5 dt5
657 - test-windows10-64/debug-mochitest-devtools-chrome-e10s-1 dt1
571 - test-windows10-64/debug-mochitest-devtools-chrome-e10s-2 dt2
557 - test-windows10-64/debug-mochitest-devtools-chrome-e10s-4 dt4
538 - test-windows10-64/debug-mochitest-media-e10s mda
498 - test-windows10-64/debug-mochitest-media-spi-e10s mda
463 - test-windows10-64/debug-mochitest-devtools-chrome-e10s-3 dt3
313 - test-windows10-64/debug-telemetry-tests-client-e10s c
275 - test-windows10-64/debug-mochitest-chrome-1proc-1 c1
156 - test-windows10-64/debug-jsreftest-e10s-1 J1
149 - test-windows10-64/debug-jsreftest-e10s-2 J2
148 - test-windows10-64/debug-mochitest-webgl1-core-e10s gl1c
138 - test-windows10-64/debug-mochitest-chrome-1proc-3 c3
119 - test-windows10-64/debug-crashtest-e10s C
115 - test-windows10-64/debug-marionette-gpu-e10s MnG
92 - test-windows10-64/debug-mochitest-gpu-e10s gpu
74 - test-windows10-64/debug-mochitest-remote-e10s remote
47 - test-windows10-64/debug-firefox-ui-functional-remote-e10s en-US
44 - test-windows10-64/debug-jsreftest-e10s-3 J3
37 - test-windows10-64/debug-mochitest-webgl2-core-e10s gl2c
28 - test-windows10-64/debug-mochitest-webgl2-ext-e10s-1 gl2e1
25 - test-windows10-64/debug-mochitest-webgl2-ext-e10s-3 gl2e3
23 - test-windows10-64/debug-mochitest-webgl2-ext-e10s-2 gl2e2
23 - test-windows10-64/debug-mochitest-webgl1-ext-e10s gl1e
20 - test-windows10-64/debug-mochitest-chrome-1proc-2 c2
19 - test-windows10-64/debug-mochitest-webgl2-ext-e10s-4 gl2e4
19 - test-windows10-64/debug-gtest-1proc GTest
6 - test-windows10-64/debug-mochitest-a11y-1proc a11y
2 - test-windows10-64/debug-xpcshell-e10s-2 X2
1 - test-windows10-64/debug-marionette-media-e10s MnM
It shows up in 17949 tests. A few of the most prevalent:
27199 - [e10s] Shutdown
1220 - [e10s] file:///Z:/task_1568804856/build/tests/reftest/tests/layout/reftests/bugs/1241631-1.html == file:///Z:/task_1568804856/build/tests/reftest/tests/layout/reftests/bugs/1241631-1-ref.html
787 - [e10s] /webdriver/tests/new_session/create_alwaysMatch.py
780 - [e10s] /webdriver/tests/new_session/create_firstMatch.py
513 - [e10s] /css/motion/offset-path-ray-004.html
512 - [e10s] /css/motion/offset-path-ray-contain-003.html
507 - [e10s] /css/motion/offset-path-ray-007.html
504 - [e10s] /css/motion/offset-path-ray-001.html
504 - [e10s] /css/motion/offset-path-ray-contain-004.html
500 - [e10s] /css/motion/offset-path-ray-009.html
Reporter | ||
Updated•5 years ago
|
Comment hidden (Intermittent Failures Robot) |
Assignee | ||
Comment 2•5 years ago
|
||
I'm going to take this, because it should be easy, and the amount of time and aggravation it will save in terms of misfiled bugs (because the IPC message is mistaken as the cause, or especially when it's interleaved with the real error and the result is misparsed by tools) should pay back the cost of doing it in maybe a week or two.
Context, which is probably already in some other bug somewhere: this is a vestige of the old idea that we'd (and I say “we” but this is before my time) have a layer of our own glue on top of the Chromium IPC classes; the signaling that the channel is being intentionally closed is in the mozilla::ipc
layer, but before we get there the Chromium code has already yelled about it on stderr. This can happen on Unix, but on Windows it's worse because we get a “broken pipe” error when reading, not only when writing.
Assignee | ||
Comment 3•5 years ago
|
||
There are two issues here:
-
These error messages occur even during normal channel shutdown,
because that's tracked in the mozilla::ipc::MessageChannel layer,
which the ipc/chromium code can't access. -
If we get this kind of error when the channel wasn't intentionally
closed, it almost certainly means that the other process crashed. In
that case, having error messages from a different process and a likely
unrelated subsystem just leads to confusion and misfiled bugs.
(Also complicating things: on Unix a closed channel often, but not
always, results in an end-of-file indication, which already isn't
logged; on Windows it's always a broken pipe error, which causes a much
larger amount of log spam.)
Bonus fix: the error that contains a fd number is clarified to avoid
having it mistaken for an error code.
Comment 5•5 years ago
|
||
bugherder |
Description
•