config/recurse.mk:93: target `intl/bidi/export' given more than once in the same rule.
Categories
(Firefox Build System :: General, defect, P3)
Tracking
(firefox125 fixed)
| Tracking | Status | |
|---|---|---|
| firefox125 | --- | fixed |
People
(Reporter: arai, Assigned: arai)
References
Details
Attachments
(4 files)
|
48 bytes,
text/x-phabricator-request
|
Details | Review | |
|
48 bytes,
text/x-phabricator-request
|
Details | Review | |
|
Bug 1880746 - Part 2: Simplify angle moz.build DIRS dependencies. r?#firefox-build-system-reviewers!
48 bytes,
text/x-phabricator-request
|
Details | Review | |
|
Bug 1880746 - Part 3: Remove duplicate DIRS from js/app.mozbuild. r?#firefox-build-system-reviewers!
48 bytes,
text/x-phabricator-request
|
Details | Review |
running ./mach build shows the following warning message
.../mozilla-unified/config/recurse.mk:93: target `intl/bidi/export' given more than once in the same rule.
This comes from intl/bidi appears in export_dirs twice in OBJDIR/root.mk, which is a kind of regression from bug 1824671 patch 11.
OBJDIR/root.mk
export_dirs := build build/rust/mozbuild config js/src js/src/shell js/src/jsapi-tests js/src/tests js/src/build js/src/frontend js/src/gc js/src/jit js/src/wasm intl/bidi memory/build memory/replace/logalloc/replay mozglue/baseprofiler mozglue/build config/external/ffi xpcom/xpidl config/external/wasm2c_sandbox_compiler media/libvpx media/libopus media/libtheora media/mp4parse-rust xpcom/idl-parser/xpidl xpcom/base xpcom/ds xpcom/components xpcom/build modules/libpref intl/bidi intl/hyphenation/glue intl/l10n intl/locale intl/unicharutil/util netwerk/base/mozurl netwerk/base/rust-helper netwerk/socket/neqo_glue ipc/app ipc/ipdl ipc/ipdl/test/ipdl testing/specialpowers gfx/webrender_bindings gfx/wgpu_bindings dom/base dom/bindings dom/bindings/test dom/fs/parent dom/media third_party/libwebrtc/experiments/registered_field_trials_header_gn dom/midi/midir_impl dom/origin-trials layout/style layout/style/test/gtest layout/generic accessible/mac accessible/base accessible/xpcom tools/profiler toolkit/components/glean toolkit/components/telemetry toolkit/components/nimbus toolkit/crashreporter toolkit/crashreporter/client toolkit/crashreporter/mozannotation_client toolkit/crashreporter/mozannotation_server toolkit/locales toolkit/mozapps/update/updater toolkit/mozapps/update/updater/updater-xpcshell toolkit/mozapps/update/tests toolkit/library/build browser browser/locales browser/themes/shared/app-marketplace-icons browser/app browser/tools/mozscreenshots/mozscreenshots/extension
The list is generated by the following line:
for tier, filter in filters:
all_dirs = self._traversal.traverse("", filter)
root_mk.add_statement("%s_dirs := %s" % (tier, " ".join(all_dirs)))
possible solution is to remove duplicate in traverse method, or after that
def traverse(self, start, filter=None):
"""
Iterate over the filtered subdirectories, following the traditional
make traversal order.
"""
| Assignee | ||
Comment 1•1 year ago
|
||
Comment 2•1 year ago
•
|
||
The real problem is that intl/bidi is recursed twice (as in, added to DIRS twice), and /that/ should be fixed.
Comment 3•1 year ago
•
|
||
Same is true of intl/components, BTW.
| Assignee | ||
Comment 4•1 year ago
|
||
Updated•1 year ago
|
| Assignee | ||
Comment 5•1 year ago
|
||
| Assignee | ||
Comment 6•1 year ago
|
||
Updated•1 year ago
|
Comment 7•1 year ago
|
||
Why not make this not a warning-worthy problem? If we can tell that it's a dupe, can't we just not re-traverse it?
The change to ANGLE makes it somewhat more fragile to changes there, since now we have "spooky action at a distance" between the generator/updater and the moz.build.
It's not a huge engineering risk to hardcode this expectation, but generally being tolerant to inputs and strict on outputs is the goal, and traversal is usually expected to be resilient to duplicate links.
| Assignee | ||
Comment 8•1 year ago
•
|
||
(In reply to Kelsey Gilbert [:jgilbert] from comment #7)
Why not make this not a warning-worthy problem? If we can tell that it's a dupe, can't we just not re-traverse it?
:glandium, can I have your opinion here?
Comment 10•1 year ago
|
||
(BTW, the warning was there, and it has been for years, which shows how effective warnings are)
Comment 11•1 year ago
|
||
Comment 12•1 year ago
|
||
| bugherder | ||
https://hg.mozilla.org/mozilla-central/rev/4c90f0d5fa7a
https://hg.mozilla.org/mozilla-central/rev/8fa3d22936fe
https://hg.mozilla.org/mozilla-central/rev/50ce64203a85
https://hg.mozilla.org/mozilla-central/rev/6031aa0ce1d7
Description
•