1717448 - Integrate and support {fmt} for c++20 std::format-style formatting in gecko

Status: RESOLVED FIXED

(Core :: XPCOM, enhancement, P3)

Description

[Nika Layzell [:nika] (ni? for response)]

In C++20, a new formatting library was defined which uses rust/python style {} format indicators for formatting strings etc. std::format also supports formatting custom types in various manners and will be expanded on in future c++ versions.

In general, gecko has continued to use printf-style formatting over stream style formatting for various reasons, including code size and style. This new format string type has improved flexibility and type safety like operator<< stream formatting, while also being performant and codesize efficient like printf.

While we cannot directly adapt c++20's implementation, as we are still using c++17 and no standard library supports the full spec yet, there is a community implementation of the syntax in the {fmt} library, which also provides extra flexibility allowing us to integrate into ns[C]String more efficiently.

This bug tracks potentially vendoring {fmt} into the tree and adding basic support to core xpcom features like MOZ_LOG and ns[C]String for the library.

Comment 1

[Mike Hommey [:glandium]]

One thing to look out for: how it formats floats (specifically, the rounding for floats)

Comment 2

[Nika Layzell [:nika] (ni? for response)]

(In reply to Mike Hommey [:glandium] from comment #1)

One thing to look out for: how it formats floats (specifically, the rounding for floats)

I looked into this a while back, and the {fmt} library appears to take glibc's approach to float rounding rather than our approach with double-conversion and ecmascript rules (from that testing I also noticed that their printf implementation appears to have some issues around padding/sign digits, however it's unclear whether this impacts their normal format strings).

If we import the crate, we'll probably want to apply some patches on the source to replace the floating point formatting infrastructure with our existing infrastructure in order to use the ecmascript rules and double-conversion under the hood, as it seems somewhat unlikely to me that they'd be interested in upstreaming ecmascript rounding support.

Comment 3

[Nika Layzell [:nika] (ni? for response)]

This proposal uses an third-party library, rather than c++20's format implementation, so doesn't depend on c++20.

Comment 4

[Paul Adenot (:padenot)]

Attachment: Bug 1717448 - Vendor {fmt} in third_party. r=glandium

Comment 5

[Paul Adenot (:padenot)]

I finally got annoyed enough.

My initial focus will be on providing better logging and assertion reporting. We can patch it to support our floating point representation in a second stage, it's around here: https://github.com/fmtlib/fmt/blob/master/include/fmt/format.h#L3589-L3612, and doesn't look horribly complicated to do or maintain (famous last words).

Comment 6

[Paul Adenot (:padenot)]

Attachment: Bug 1717448 - Teach MOZ_LOG to use {fmt}. r=nika

Comment 7

[Paul Adenot (:padenot)]

Attachment: Bug 1717448 - Print the diagnostic message in AssertedCast with {fmt}. r=nbp

Comment 8

[Paul Adenot (:padenot)]

Attachment: Bug 1717448 - Replace printf-style logging by MOZ_LOG_FMT in AudioData.cpp. r=nika

Comment 9

[Paul Adenot (:padenot)]

Attachment: WIP: Bug 1717448 - DO NOT LAND Example on how to quickly do a debug printout using {fmt}.

Comment 10

[Paul Adenot (:padenot)]

Attachment: Bug 1717448 - Add nsFmtCString, and ns*String::AppendFmt. r?nika

Comment 11

[Paul Adenot (:padenot)]

Attachment: Bug 1717448 - Allow formatting double like EcmaScript expects using {fmt}. r?nika

Comment 12

[Paul Adenot (:padenot)]

Attachment: Bug 1717448 - Add some tests for Gecko integration. r=glandium

Comment 13

[Paul Adenot (:padenot)]

Attachment: Bug 1717448 - Add some documentation about using {fmt} in Gecko. r=nika

Comment 14

[Paul Adenot (:padenot)]

This seems to be useful with this initial set of patches, but happy to hear what people would like added. I'm thinking of writing a big bag of regexp to ease conversion.

Comment 15

[Nika Layzell [:nika] (ni? for response)]

Attachment: Old patch from 2022 for nsCString integration

Attaching an old patch with my WIP changes from when I last looked into this in 2022 where I was adding nsCString support for {fmt}.

Comment 16

[Nika Layzell [:nika] (ni? for response)]

Also FWIW, I believe we are much closer to building with c++20 than we were 2-3 years ago when I first looked into this, which might mean that holding out for std::format might make more sense in terms of keeping dependencies down, though I suppose std::format is unlikely to do things like rounding of doubles in the way that we want it to.

It might be worth waiting for c++20 before trying to do this so we can properly compare the options we have available.

(In reply to Paul Adenot (:padenot) from comment #14)

This seems to be useful with this initial set of patches, but happy to hear what people would like added. I'm thinking of writing a big bag of regexp to ease conversion.

Given the downsides of {fmt} (like not checking format strings at compile time reliably compared to printf, crashing on errors without options for recovery, and probably having a larger binary size overhead in the non-formatting cases), we may want to hold off from trying to do any sort of bulk transition to {fmt}.

Comment 17

[Paul Adenot (:padenot)]

(In reply to Nika Layzell [:nika] (ni? for response) from comment #16)

Also FWIW, I believe we are much closer to building with c++20 than we were 2-3 years ago when I first looked into this

My understanding is that this is unfortunately still years away, and that was what prompted me to do this. https://bugzilla.mozilla.org/show_bug.cgi?id=1880779#c2 has some info, and if I read :glandium's messages correctly, https://glandium.github.io/firefox-linux-compat-matrix/ has ballpark numbers (maybe not exactly correct, but not horribly incorrect). We can't do C++20 until we don't support RHEL8, and that is in a long time. Mike, please correct me if I understood things incorrectly.

Given the downsides of {fmt} (like not checking format strings at compile time reliably compared to printf, crashing on errors without options for recovery, and probably having a larger binary size overhead in the non-formatting cases), we may want to hold off from trying to do any sort of bulk transition to {fmt}.

Indeed, but I've been wrong more than once in terms of binary size, so I'm planning to do a conversion on a few of my modules to get a sense of it. My patch set uses type erasure to keep the binary size down, and we can use compile-time checking in lots of cases.

Comment 18

[Paul Adenot (:padenot)]

Attachment: Bug 1717448 - Enforce the use of FMT_STRING at call sites. r?nika,glandium

Comment 19

[Paul Adenot (:padenot)]

Attachment: Bug 1717448 - Integrate {fmt} with nsString. r=#xpcom-reviewers,mccr8,nika

Comment 20

[Paul Adenot (:padenot)]

Attachment: Bug 1717448 - Patch `{fmt}` to avoid crashing when printing `nullptr` strings, and attempting to print to an invalid FILE*. r=nika

This brings {fmt} closer to our printf.

Comment 21

[Paul Adenot (:padenot)]

Attachment: Bug 1717448 - Patch fmt to qualify back_inserter to fix the build. r?nika

Comment 22

[Paul Adenot (:padenot)]

Attachment: Bug 1717448 - Make `{fmt}` use `double-conversion` for converting float to string internally. r=glandium,nika

Comment 23

[Nika Layzell [:nika] (ni? for response)]

Attachment: Old patch from 2022 for testing {fmt} rounding behaviour

Comment 24

[Mike Hommey [:glandium]]

(In reply to Paul Adenot (:padenot) from comment #17)

My understanding is that this is unfortunately still years away, and that was what prompted me to do this. https://bugzilla.mozilla.org/show_bug.cgi?id=1880779#c2 has some info, and if I read :glandium's messages correctly, https://glandium.github.io/firefox-linux-compat-matrix/ has ballpark numbers (maybe not exactly correct, but not horribly incorrect). We can't do C++20 until we don't support RHEL8, and that is in a long time. Mike, please correct me if I understood things incorrectly.

The situation is a little bit more complicated, but as a super simplified version, yes. Where the complication enters, is that RHEL updates do add some new compilers, which may allow us to bump our baseline, BUT, full C++20 support also depends on libstdc++, although depending on the bits we use, that might not be a problem. I don't know if std::format requires new symbols from libstd++ (but that's likely).

Comment 25

[Paul Adenot (:padenot)]

Attachment: Bug 1717448 - Exclude {fmt} from gecko-specific checks. r=glandium,nika

It looks like the excluding logic around ThirdPartyPaths.txt works on include
paths, and not on the path of files on disk, so it's necessary to explicitly
exclude fmt/base.h and fmt/format.h from two checks that fail as things
stand today:

• the one that checks that explicit is used for single-arg ctors, this is used
  throughout {fmt}
• the one that checks that val != val isn't used to check for NaN. This is
  used in {fmt} because std::isnan doesn't have an overload for __float128.

Comment 26

[Paul Adenot (:padenot)]

Attachment: Bug 1717448 - Don't use std::generic_category and other error reporting facilities, ignore IO errors. r=glandium

Comment 27

[Paul Adenot (:padenot)]

Attachment: Bug 1717448 - Ignore allocations in done within `{fmt}`. r=sfink

Comment 28

[Paul Adenot (:padenot)]

https://treeherder.mozilla.org/jobs?repo=try&revision=de2071141a052bc5a83b87f9c2b09e4c894b2b5b
https://treeherder.mozilla.org/jobs?repo=try&revision=f99bfb9150db9a97b3729fdc5a204242246045a1

looks alright, not sure what's up with the H job though.

Comment 29

[Mike Hommey [:glandium]]

(In reply to Paul Adenot (:padenot) from comment #28)

https://treeherder.mozilla.org/jobs?repo=try&revision=de2071141a052bc5a83b87f9c2b09e4c894b2b5b
https://treeherder.mozilla.org/jobs?repo=try&revision=f99bfb9150db9a97b3729fdc5a204242246045a1

looks alright, not sure what's up with the H job though.

Base toolchain tasks are not happy.

Comment 30

[Mike Hommey [:glandium]]

For the H job, try asking @sfink.

Comment 31

[Paul Adenot (:padenot)]

Attachment: Bug 1717448 - Exclude `{fmt}` internals from hazard analysis. r=sfink

Apparently {fmt} uses a function pointer under the hood and that trips our
analysis. In practice this excluded call-site will only print numbers (of
various width and signedness, and also floats), in debug, when the assertion
fails.

Comment 32

[Paul Adenot (:padenot)]

Attachment: Bug 1717448 - Use glibc's printf test suite for double formatting to cross-check various ways of formatting doubles. r?glandium

Comment 33

[Paul Adenot (:padenot)]

Attachment: Bug 1717448 - Reinstate a type_identity utility, removed by {fmt} upstream but needed. r?nika

Comment 34

[Paul Adenot (:padenot)]

I've updated all patches here, with current upstream tip + rebased all patches + address all comments except mccr8's about asserting in the nsString adapter, that I will have a look at now.

Notable changes:

• Our injected float formatting code is cross-checked with Gecko's Printf.cpp implementation, using glibc's test suite, including with translating format string from printf to fmt, bringing more confidence that all this works
• we can drop the patch that was adding a using namespace std;, fixed upstream
• merged two patches about error reporting / IO error (content unchanged)
• had to add our own type_identity util, this is C++20, but isn't upstream any more
• fixed padding handling when formatting floating point, in some cases we could intermix the padding and the sign (e.g. 000-123.123 would be output instead of -000123.123, but -123.123 is also correctly output: 0s are after the sign, other chars before the sign).
• needed a couple more header files to compile (e.g. ostream.h)

The rest is largely the same modulo trivial rebases.

Comment 35

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/integration/autoland/rev/adef9b3b6231
Vendor {fmt} in third_party. r=glandium
https://hg.mozilla.org/integration/autoland/rev/47ed151719a4
Patch `{fmt}` to avoid crashing when printing `nullptr` strings, and attempting to print to an invalid FILE*. r=nika
https://hg.mozilla.org/integration/autoland/rev/2e17fb168eaf
Make `{fmt}` use `double-conversion` for converting float to string internally. r=glandium,nika
https://hg.mozilla.org/integration/autoland/rev/993a64fd729a
Don't use std::generic_category and other error reporting facilities, ignore IO errors. r=glandium
https://hg.mozilla.org/integration/autoland/rev/f0ca736c051b
Exclude `{fmt}` internals from hazard analysis. r=sfink
https://hg.mozilla.org/integration/autoland/rev/4381f24f8495
Exclude {fmt} from gecko-specific checks. r=glandium,nika
https://hg.mozilla.org/integration/autoland/rev/aebf794911ca
Ignore allocations in done within `{fmt}`. r=glandium
https://hg.mozilla.org/integration/autoland/rev/8295836348fb
Teach MOZ_LOG to use {fmt}. r=glandium,nika
https://hg.mozilla.org/integration/autoland/rev/5e5b42a7cde1
Print the diagnostic message in AssertedCast with {fmt}. r=nbp
https://hg.mozilla.org/integration/autoland/rev/37ab78b00a4f
Integrate {fmt} with nsString. r=#xpcom-reviewers,nika
https://hg.mozilla.org/integration/autoland/rev/670256c85479
Add some tests for Gecko integration. r=nika,glandium
https://hg.mozilla.org/integration/autoland/rev/41c8a260d55b
Add some documentation about using {fmt} in Gecko. r=nika
https://hg.mozilla.org/integration/autoland/rev/4254666ea8a7
Replace printf-style logging by MOZ_LOG_FMT in AudioData.cpp. r=nika

Comment 36

[tszentpeteri]

Backed out for causing gtest failures on iofwrite.c

• Backout link
• Push with failures
• Failure Log:

[task 2024-10-18T16:31:41.153Z] 16:31:41     INFO -  TEST-START | Fmt.IOError
[task 2024-10-18T16:31:41.153Z] 16:31:41     INFO -  AddressSanitizer:DEADLYSIGNAL
[task 2024-10-18T16:31:41.153Z] 16:31:41     INFO -  =================================================================
[task 2024-10-18T16:31:41.154Z] 16:31:41    ERROR -  ==992==ERROR: AddressSanitizer: SEGV on unknown address (pc 0x7f94901e6950 bp 0x000000000006 sp 0x7fffbb4cdcd0 T0)
[task 2024-10-18T16:31:41.154Z] 16:31:41     INFO -  ==992==The signal is caused by a READ memory access.
[task 2024-10-18T16:31:41.154Z] 16:31:41     INFO -  ==992==Hint: this fault was caused by a dereference of a high value address (see register values below).  Disassemble the provided pc to learn which register was used.
[task 2024-10-18T16:31:41.697Z] 16:31:41     INFO -      #0 0x7f94901e6950 in _IO_fwrite /tmp/glibc/libio/iofwrite.c:37
[task 2024-10-18T16:31:41.697Z] 16:31:41     INFO -      #1 0x55622dc91fa9 in fwrite /builds/worker/fetches/llvm-project/compiler-rt/lib/asan/../sanitizer_common/sanitizer_common_interceptors.inc
[task 2024-10-18T16:31:41.698Z] 16:31:41     INFO -      #2 0x55622de24618 in fwrite_all /builds/worker/workspace/obj-build/dist/include/fmt/format-inl.h:77:3
[task 2024-10-18T16:31:41.698Z] 16:31:41     INFO -      #3 0x55622de24618 in print /builds/worker/workspace/obj-build/dist/include/fmt/format-inl.h:1727:3
[task 2024-10-18T16:31:41.698Z] 16:31:41     INFO -      #4 0x55622de24618 in fmt::v11::vprintln(_IO_FILE*, fmt::v11::basic_string_view<char>, fmt::v11::basic_format_args<fmt::v11::context>) /builds/worker/workspace/obj-build/dist/include/fmt/format-inl.h:1748:3
[task 2024-10-18T16:31:41.698Z] 16:31:41     INFO -      #5 0x7f946a058018 in println<int> /builds/worker/workspace/obj-build/dist/include/fmt/base.h:2887:29
[task 2024-10-18T16:31:41.698Z] 16:31:41     INFO -      #6 0x7f946a058018 in Fmt_IOError_Test::TestBody() /builds/worker/checkouts/gecko/mozglue/tests/gtest/TestFmt.cpp:208:3
[task 2024-10-18T16:31:41.698Z] 16:31:41     INFO -      #7 0x7f946b4b8191 in testing::Test::Run() /builds/worker/checkouts/gecko/third_party/googletest/googletest/src/gtest.cc:2713:5
[task 2024-10-18T16:31:41.698Z] 16:31:41     INFO -      #8 0x7f946b4ba5a9 in testing::TestInfo::Run() /builds/worker/checkouts/gecko/third_party/googletest/googletest/src/gtest.cc:2859:11
[task 2024-10-18T16:31:41.699Z] 16:31:41     INFO -      #9 0x7f946b4bc9e2 in testing::TestSuite::Run() /builds/worker/checkouts/gecko/third_party/googletest/googletest/src/gtest.cc:3037:30
[task 2024-10-18T16:31:41.699Z] 16:31:41     INFO -      #10 0x7f946b4e9687 in testing::internal::UnitTestImpl::RunAllTests() /builds/worker/checkouts/gecko/third_party/googletest/googletest/src/gtest.cc:5967:44
[task 2024-10-18T16:31:41.699Z] 16:31:41     INFO -      #11 0x7f946b4e888b in HandleExceptionsInMethodIfSupported<testing::internal::UnitTestImpl, bool> /builds/worker/checkouts/gecko/third_party/googletest/googletest/src/gtest.cc
[task 2024-10-18T16:31:41.699Z] 16:31:41     INFO -      #12 0x7f946b4e888b in testing::UnitTest::Run() /builds/worker/checkouts/gecko/third_party/googletest/googletest/src/gtest.cc:5546:10
[task 2024-10-18T16:31:41.699Z] 16:31:41     INFO -      #13 0x7f946b479f1f in RUN_ALL_TESTS /builds/worker/workspace/obj-build/dist/include/gtest/gtest.h:2336:73
[task 2024-10-18T16:31:41.700Z] 16:31:41     INFO -      #14 0x7f946b479f1f in mozilla::RunGTestFunc(int*, char**) /builds/worker/checkouts/gecko/testing/gtest/mozilla/GTestRunner.cpp:172:10
[task 2024-10-18T16:31:41.700Z] 16:31:41     INFO -      #15 0x7f947990cd1e in XREMain::XRE_mainStartup(bool*) /builds/worker/checkouts/gecko/toolkit/xre/nsAppRunner.cpp:4716:16
[task 2024-10-18T16:31:41.700Z] 16:31:41     INFO -      #16 0x7f947991842f in XREMain::XRE_main(int, char**, mozilla::BootstrapConfig const&) /builds/worker/checkouts/gecko/toolkit/xre/nsAppRunner.cpp:6007:12
[task 2024-10-18T16:31:41.701Z] 16:31:41     INFO -      #17 0x7f94799195d3 in XRE_main(int, char**, mozilla::BootstrapConfig const&) /builds/worker/checkouts/gecko/toolkit/xre/nsAppRunner.cpp:6105:21
[task 2024-10-18T16:31:41.701Z] 16:31:41     INFO -      #18 0x55622dd4c17d in do_main /builds/worker/checkouts/gecko/browser/app/nsBrowserApp.cpp:233:22
[task 2024-10-18T16:31:41.701Z] 16:31:41     INFO -      #19 0x55622dd4c17d in main /builds/worker/checkouts/gecko/browser/app/nsBrowserApp.cpp:465:16
[task 2024-10-18T16:31:41.701Z] 16:31:41     INFO -      #20 0x7f9490188b96 in __libc_start_main /tmp/glibc/csu/../csu/libc-start.c:310
[task 2024-10-18T16:31:41.701Z] 16:31:41     INFO -      #21 0x55622dc73c58 in _start (/builds/worker/workspace/build/application/firefox/firefox+0xc8c58) (BuildId: d2c4a70f6b8c7ca87da0a6df9aba41a74a4798f4)
[task 2024-10-18T16:31:41.702Z] 16:31:41     INFO -  AddressSanitizer can not provide additional info.
[task 2024-10-18T16:31:41.702Z] 16:31:41     INFO -  SUMMARY: AddressSanitizer: SEGV /tmp/glibc/libio/iofwrite.c:37 in _IO_fwrite
[task 2024-10-18T16:31:41.702Z] 16:31:41     INFO -  ==992==ABORTING
[task 2024-10-18T16:31:41.716Z] 16:31:41     INFO -  gtest INFO | gtest | process wait complete, returncode=1
[task 2024-10-18T16:31:41.718Z] 16:31:41     INFO -  mozcrash checking /builds/worker/workspace/build/tests/gtest for minidumps...
[task 2024-10-18T16:31:41.718Z] 16:31:41  WARNING -  gtest TEST-UNEXPECTED-FAIL | gtest | test failed with return code 1
[task 2024-10-18T16:31:41.718Z] 16:31:41     INFO -  gtest INFO | rungtests.py exits with code 1
[task 2024-10-18T16:31:41.739Z] 16:31:41     INFO - Return code: 1
[task 2024-10-18T16:31:41.739Z] 16:31:41    ERROR - No tests run or test summary not found
[task 2024-10-18T16:31:41.739Z] 16:31:41     INFO - TinderboxPrint: gtest-gtest<br/><em class="testfail">T-FAIL</em>
[task 2024-10-18T16:31:41.739Z] 16:31:41  WARNING - setting return code to 2
[task 2024-10-18T16:31:41.739Z] 16:31:41     INFO - The gtest suite: gtest ran with return status: FAILURE
[task 2024-10-18T16:31:41.739Z] 16:31:41     INFO - Running post-action listener: _package_coverage_data
[task 2024-10-18T16:31:41.739Z] 16:31:41     INFO - Running post-action listener: _resource_record_post_action
[task 2024-10-18T16:31:41.739Z] 16:31:41     INFO - Running post-action listener: process_java_coverage_data
[task 2024-10-18T16:31:41.740Z] 16:31:41     INFO - [mozharness: 2024-10-18 16:31:41.739941Z] Finished run-tests step (success)
[task 2024-10-18T16:31:41.740Z] 16:31:41     INFO - [mozharness: 2024-10-18 16:31:41.740094Z] Running uninstall step.
[task 2024-10-18T16:31:41.740Z] 16:31:41     INFO - Running pre-action listener: _resource_record_pre_action
[task 2024-10-18T16:31:41.740Z] 16:31:41     INFO - Running main action method: uninstall
[task 2024-10-18T16:31:41.740Z] 16:31:41     INFO - Skipping uninstall for non-MSIX test
[task 2024-10-18T16:31:41.740Z] 16:31:41     INFO - Running post-action listener: _resource_record_post_action
[task 2024-10-18T16:31:41.740Z] 16:31:41     INFO - [mozharness: 2024-10-18 16:31:41.740373Z] Finished uninstall step (success)
[task 2024-10-18T16:31:41.740Z] 16:31:41     INFO - Running post-run listener: _resource_record_post_run
[task 2024-10-18T16:31:41.803Z] 16:31:41     INFO - Validating Perfherder data against /builds/worker/workspace/mozharness/external_tools/performance-artifact-schema.json
[task 2024-10-18T16:31:41.805Z] 16:31:41     INFO - PERFHERDER_DATA: {"framework": {"name": "job_resource_usage"}, "suites": [{"name": "gtest.gtest.overall", "extraOptions": ["taskcluster-projects/887720501152/machineTypes/c2-standard-4"], "subtests": [{"name": "cpu_percent", "value": 22.766557971014475}, {"name": "io_write_bytes", "value": 2829402112}, {"name": "io.read_bytes", "value": 4096}, {"name": "io_write_time", "value": 3124612}, {"name": "io_read_time", "value": 228}]}, {"name": "gtest.gtest.start-pulseaudio", "subtests": [{"name": "time", "value": 0.01823795300015263}, {"name": "cpu_percent", "value": 0}]}, {"name": "gtest.gtest.install", "subtests": [{"name": "time", "value": 57.822186512999906}, {"name": "cpu_percent", "value": 25.276906412478315}]}, {"name": "gtest.gtest.stage-files", "subtests": [{"name": "time", "value": 1.7499074580000524}, {"name": "cpu_percent", "value": 16.371875}]}, {"name": "gtest.gtest.run-tests", "subtests": [{"name": "time", "value": 9.418066888999874}, {"name": "cpu_percent", "value": 8.565425531914894}]}, {"name": "gtest.gtest.uninstall", "subtests": [{"name": "time", "value": 0.00016078400039987173}, {"name": "cpu_percent", "value": 0}]}]}
[task 2024-10-18T16:31:41.805Z] 16:31:41     INFO - Total resource usage - Wall time: 69s; CPU: Can't collect data; Read bytes: 4096; Write bytes: 2829402112; Read time: 228; Write time: 3124612
[task 2024-10-18T16:31:41.805Z] 16:31:41     INFO - TinderboxPrint: I/O read bytes / time<br/>4,096 / 228
[task 2024-10-18T16:31:41.805Z] 16:31:41     INFO - TinderboxPrint: I/O write bytes / time<br/>2,829,402,112 / 3,124,612
[task 2024-10-18T16:31:41.805Z] 16:31:41     INFO - TinderboxPrint: CPU idle<br/>188.4 (68.4%)
[task 2024-10-18T16:31:41.805Z] 16:31:41     INFO - TinderboxPrint: CPU iowait<br/>24.2 (8.8%)
[task 2024-10-18T16:31:41.805Z] 16:31:41     INFO - TinderboxPrint: CPU system<br/>3.8 (1.4%)
[task 2024-10-18T16:31:41.805Z] 16:31:41     INFO - TinderboxPrint: CPU user<br/>59.0 (21.4%)
[task 2024-10-18T16:31:41.805Z] 16:31:41     INFO - TinderboxPrint: Swap in / out<br/>0 / 0
[task 2024-10-18T16:31:41.805Z] 16:31:41     INFO - start-pulseaudio - Wall time: 0s; CPU: Can't collect data; Read bytes: 0; Write bytes: 0; Read time: 0; Write time: 0
[task 2024-10-18T16:31:41.808Z] 16:31:41     INFO - install - Wall time: 58s; CPU: 25%; Read bytes: 4096; Write bytes: 1728339968; Read time: 228; Write time: 1890776
[task 2024-10-18T16:31:41.809Z] 16:31:41     INFO - stage-files - Wall time: 2s; CPU: 16%; Read bytes: 0; Write bytes: 151699456; Read time: 0; Write time: 127776
[task 2024-10-18T16:31:41.810Z] 16:31:41     INFO - run-tests - Wall time: 9s; CPU: 9%; Read bytes: 0; Write bytes: 935993344; Read time: 0; Write time: 1091392
[task 2024-10-18T16:31:41.810Z] 16:31:41     INFO - uninstall - Wall time: 0s; CPU: Can't collect data; Read bytes: 0; Write bytes: 0; Read time: 0; Write time: 0
[task 2024-10-18T16:31:41.926Z] 16:31:41  WARNING - returning nonzero exit status 2
[task 2024-10-18T16:31:41.952Z] cleanup
[task 2024-10-18T16:31:41.952Z] + cleanup
[task 2024-10-18T16:31:41.952Z] + local rv=2
[task 2024-10-18T16:31:41.952Z] + [[ -s /builds/worker/.xsession-errors ]]
[task 2024-10-18T16:31:41.952Z] + cp /builds/worker/.xsession-errors /builds/worker/artifacts/public/xsession-errors.log
[task 2024-10-18T16:31:41.954Z] + '[' ']'
[task 2024-10-18T16:31:41.954Z] + true
[task 2024-10-18T16:31:41.954Z] + cleanup_xvfb
[task 2024-10-18T16:31:41.954Z] ++ pidof Xvfb
[task 2024-10-18T16:31:41.957Z] + local xvfb_pid=50
[task 2024-10-18T16:31:41.957Z] + local vnc=false
[task 2024-10-18T16:31:41.957Z] + local interactive=false
[task 2024-10-18T16:31:41.957Z] + '[' -n 50 ']'
[task 2024-10-18T16:31:41.957Z] + [[ false == false ]]
[task 2024-10-18T16:31:41.957Z] + [[ false == false ]]
[task 2024-10-18T16:31:41.957Z] + kill 50
[task 2024-10-18T16:31:41.957Z] + screen -XS xvfb quit
[task 2024-10-18T16:31:41.976Z] + exit 2
[taskcluster 2024-10-18 16:31:42.558Z] === Task Finished ===
[taskcluster 2024-10-18 16:31:43.571Z] Unsuccessful task run with exit code: 2 completed in 173.889 seconds

Comment 37

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/integration/autoland/rev/9719b35b85d6
Vendor {fmt} in third_party. r=glandium
https://hg.mozilla.org/integration/autoland/rev/e595c59c8b0c
Patch `{fmt}` to avoid crashing when printing `nullptr` strings, and attempting to print to an invalid FILE*. r=nika
https://hg.mozilla.org/integration/autoland/rev/a6a266e1cedf
Make `{fmt}` use `double-conversion` for converting float to string internally. r=glandium,nika
https://hg.mozilla.org/integration/autoland/rev/f2c1cb4d8185
Don't use std::generic_category and other error reporting facilities, ignore IO errors. r=glandium
https://hg.mozilla.org/integration/autoland/rev/54fd0897fac3
Exclude `{fmt}` internals from hazard analysis. r=sfink
https://hg.mozilla.org/integration/autoland/rev/7fa5b359b4c3
Exclude {fmt} from gecko-specific checks. r=glandium,nika
https://hg.mozilla.org/integration/autoland/rev/a4adf3ab7665
Ignore allocations in done within `{fmt}`. r=sfink
https://hg.mozilla.org/integration/autoland/rev/683a40d698a3
Teach MOZ_LOG to use {fmt}. r=nika
https://hg.mozilla.org/integration/autoland/rev/9bd358abca26
Print the diagnostic message in AssertedCast with {fmt}. r=nbp
https://hg.mozilla.org/integration/autoland/rev/ffa33e073c36
Integrate {fmt} with nsString. r=xpcom-reviewers,nika,mccr8
https://hg.mozilla.org/integration/autoland/rev/9b55c0eea2be
Add some tests for Gecko integration. r=glandium
https://hg.mozilla.org/integration/autoland/rev/3ada13e9d071
Add some documentation about using {fmt} in Gecko. r=nika
https://hg.mozilla.org/integration/autoland/rev/e873c27d97a9
Replace printf-style logging by MOZ_LOG_FMT in AudioData.cpp. r=nika
https://hg.mozilla.org/integration/autoland/rev/2548732ae2f3
apply code formatting via Lando

Comment 38

[Norisz Fay [:noriszfay]]

https://hg.mozilla.org/mozilla-central/rev/9719b35b85d6
https://hg.mozilla.org/mozilla-central/rev/e595c59c8b0c
https://hg.mozilla.org/mozilla-central/rev/a6a266e1cedf
https://hg.mozilla.org/mozilla-central/rev/f2c1cb4d8185
https://hg.mozilla.org/mozilla-central/rev/54fd0897fac3
https://hg.mozilla.org/mozilla-central/rev/7fa5b359b4c3
https://hg.mozilla.org/mozilla-central/rev/a4adf3ab7665
https://hg.mozilla.org/mozilla-central/rev/683a40d698a3
https://hg.mozilla.org/mozilla-central/rev/9bd358abca26
https://hg.mozilla.org/mozilla-central/rev/ffa33e073c36
https://hg.mozilla.org/mozilla-central/rev/9b55c0eea2be
https://hg.mozilla.org/mozilla-central/rev/3ada13e9d071
https://hg.mozilla.org/mozilla-central/rev/e873c27d97a9
https://hg.mozilla.org/mozilla-central/rev/2548732ae2f3