multiple definitions of AUTO_PROFILER_THREAD_WAKE on non-MOZ_GECKO_PROFILER platforms
Categories
(Core :: Gecko Profiler, defect, P3)
Tracking
()
Tracking | Status | |
---|---|---|
firefox-esr91 | --- | unaffected |
firefox100 | --- | wontfix |
firefox101 | --- | wontfix |
firefox102 | --- | fixed |
firefox103 | --- | fixed |
People
(Reporter: gaston, Assigned: mozbugz)
References
(Regression)
Details
(Keywords: regression)
Attachments
(1 file)
Bug 1770544 - Avoid AUTO_PROFILER_THREAD_WAKE re-#define in non-MOZ_GECKO_PROFILER builds - r?gaston
48 bytes,
text/x-phabricator-request
|
pascalc
:
approval-mozilla-beta+
|
Details | Review |
this one is filling my build logs since the profiler headers are included from everywhere:
7:23.02 /usr/obj/m-c/dist/include/mozilla/ProfilerThreadState.h:90:9: warning: 'AUTO_PROFILER_THREAD_WAKE' macro redefined [-Wmacro-redefined]
7:23.02 #define AUTO_PROFILER_THREAD_WAKE mozilla::AutoProfilerThreadWake PROFILER_RAII
7:23.02 ^
7:23.02 /usr/obj/m-c/dist/include/mozilla/ProfilerThreadSleep.h:21:11: note: previous definition is here
7:23.03 # define AUTO_PROFILER_THREAD_WAKE
maybe an #ifndef could be added to the fallback definition in ProfilerThreadSleep.h ?
Assignee | ||
Comment 1•3 years ago
|
||
Thank you for the report.
Hmm, the one in ProfilerThreadSleep.h is an empty fallback for non-MOZ_GECKO_PROFILER
builds, but the one in ProfilerThradState.h is always defined.
I rather think that the "real" definition in ProfilerThreadState.h should be guarded with in #ifdef MOZ_GECKO_PROFILER
, similar to how AUTO_PROFILER_THREAD_SLEEP
is.
I may not have the time for it this week (traveling). Keeping the NI as a reminder.
Comment 2•3 years ago
|
||
Set release status flags based on info from the regressing bug 1757596
Reporter | ||
Comment 3•3 years ago
|
||
(In reply to Gerald Squelart [:gerald] (he/him) from comment #1)
Thank you for the report.
Hmm, the one in ProfilerThreadSleep.h is an empty fallback for non-MOZ_GECKO_PROFILER
builds, but the one in ProfilerThradState.h is always defined.I rather think that the "real" definition in ProfilerThreadState.h should be guarded with in
#ifdef MOZ_GECKO_PROFILER
, similar to howAUTO_PROFILER_THREAD_SLEEP
is.
you mean adding #ifdef MOZ_GECKO_PROFILER
around https://searchfox.org/mozilla-central/source/tools/profiler/public/ProfilerThreadState.h#90 ? or also include AutoProfilerThreadWake class definition within it ? the class is also defined here: https://searchfox.org/mozilla-central/source/mozglue/baseprofiler/public/BaseProfiler.h#467
I may not have the time for it this week (traveling). Keeping the NI as a reminder.
No hurry, can test patch when you have one to confirm the warnings are gone on platforms without profiler :)
Updated•3 years ago
|
Updated•3 years ago
|
Comment 4•3 years ago
|
||
Set release status flags based on info from the regressing bug 1757596
Assignee | ||
Updated•3 years ago
|
Assignee | ||
Comment 5•3 years ago
|
||
Comment 6•3 years ago
|
||
The severity field is not set for this bug.
:gerald, could you have a look please?
For more information, please visit auto_nag documentation.
Assignee | ||
Updated•3 years ago
|
Comment 8•3 years ago
|
||
bugherder |
Reporter | ||
Comment 9•3 years ago
|
||
Comment on attachment 9280151 [details]
Bug 1770544 - Avoid AUTO_PROFILER_THREAD_WAKE re-#define in non-MOZ_GECKO_PROFILER builds - r?gaston
Beta/Release Uplift Approval Request
- User impact if declined: annoying warnings during on tier3 / non-profiler platforms - would be nice to have in 102esr branch.
- Is this code covered by automated tests?: No
- Has the fix been verified in Nightly?: Yes
- Needs manual test from QE?: No
- If yes, steps to reproduce:
- List of other uplifts needed: None
- Risk to taking this patch: Low
- Why is the change risky/not risky? (and alternatives if risky): npotb for tier1 via #ifdefs..
- String changes made/needed: none
- Is Android affected?: No
ESR Uplift Approval Request
- If this is not a sec:{high,crit} bug, please state case for ESR consideration: would be nice to have in 102esr branch for tidier logs.
- User impact if declined: annoying warnings during on tier3 / non-profiler platforms
- Fix Landed on Version: 103
- Risk to taking this patch: Low
- Why is the change risky/not risky? (and alternatives if risky): npotb for tier1 via #ifdefs..
Comment 10•3 years ago
|
||
Comment on attachment 9280151 [details]
Bug 1770544 - Avoid AUTO_PROFILER_THREAD_WAKE re-#define in non-MOZ_GECKO_PROFILER builds - r?gaston
Approved for 102 beta 6, thanks.
Comment 11•3 years ago
|
||
bugherder uplift |
Updated•3 years ago
|
Updated•3 years ago
|
Description
•