Closed Bug 558518 Opened 14 years ago Closed 14 years ago

Investigate and sync-up remaining diffs from rules.mk from c-c to m-c

Categories

(MailNews Core :: Build Config, defect)

defect
Not set
minor

Tracking

(Not tracked)

RESOLVED FIXED

People

(Reporter: Callek, Assigned: Callek)

References

Details

Attachments

(2 files, 5 obsolete files)

It dawns on me that we should not only continue to track http://dev.seamonkey.at/?d=x&i=mozilla&m=c but perform a brief audit on our build-system sync like this.

I'll also do a 1.9.2 sync-check if I finish before c-c branches.
I'll just be skimming through this over the next week, citing bugs, _and_ citing changes we want to keep; I'll ignore any MOZ_1_9_2 ifdef branches here [for now]
Comment on attachment 438233 [details] [diff] [review]
m-c vs c-c as of today. |- == m-c-only|  |+ == c-c-only|

>--- mozilla/config/rules.mk	2010-04-08 23:06:06 -0400
>+++ config/rules.mk	2010-04-09 23:28:51 -0400
> ifndef MOZILLA_DIR
>-MOZILLA_DIR = $(topsrcdir)
>+MOZILLA_DIR = $(MOZILLA_SRCDIR)
> endif

Intended, (and needed)

>-testxpcobjdir = $(DEPTH)/_tests/xpcshell
>+testxpcobjdir = $(MOZDEPTH)/_tests/xpcshell

>-TEST_INSTALLER = $(PYTHON) $(topsrcdir)/config/nsinstall.py
>+TEST_INSTALLER = $(PYTHON) $(MOZILLA_SRCDIR)/config/nsinstall.py

Also intended.

> 	$(PYTHON) $(MOZILLA_DIR)/config/buildlist.py \
>-          $(testxpcobjdir)/all-test-dirs.list \
>-          $(addprefix $(MODULE)/,$(XPCSHELL_TESTS))
>+	  $(testxpcobjdir)/all-test-dirs.list \
>+	  $(addprefix $(MODULE)/,$(XPCSHELL_TESTS))

Whitespace: TODO

>-testxpcsrcdir = $(topsrcdir)/testing/xpcshell
>+testxpcsrcdir = $(MOZILLA_SRCDIR)/testing/xpcshell

Intended

> # Execute all tests in the $(XPCSHELL_TESTS) directories.
>-# See also testsuite-targets.mk 'xpcshell-tests' target for global execution.
>+# See also $(MOZILLA_DIR)/testing/testsuite-targets.mk 'xpcshell-tests' target for global execution.

Comment -> Correct and intended

> xpcshell-tests:
>-	$(PYTHON) -u $(topsrcdir)/config/pythonpath.py \
>-          -I$(topsrcdir)/build \
>+	$(PYTHON) -u $(MOZILLA_DIR)/config/pythonpath.py \
>+	  -I$(MOZILLA_DIR)/build \

Intended... and...

>-          $(testxpcsrcdir)/runxpcshelltests.py \
>-          --symbols-path=$(DIST)/crashreporter-symbols \
>-          $(DIST)/bin/xpcshell \
>-          $(foreach dir,$(XPCSHELL_TESTS),$(testxpcobjdir)/$(MODULE)/$(dir))
>+	  $(testxpcsrcdir)/runxpcshelltests.py \
>+	  --symbols-path=$(DIST)/crashreporter-symbols \
>+	  $(DIST)/bin/xpcshell \
>+	  $(foreach dir,$(XPCSHELL_TESTS),$(testxpcobjdir)/$(MODULE)/$(dir))

Whitespace: TODO

> check-interactive:
>-	$(PYTHON) -u $(topsrcdir)/config/pythonpath.py \
>-          -I$(topsrcdir)/build \
>+	$(PYTHON) -u $(MOZILLA_DIR)/config/pythonpath.py \
>+	  -I$(MOZILLA_DIR)/build \

Intended... and

>-          $(testxpcsrcdir)/runxpcshelltests.py \
>-          --symbols-path=$(DIST)/crashreporter-symbols \
>-          --test-path=$(SOLO_FILE) \
>-          --interactive \
>-          $(DIST)/bin/xpcshell \
>-          $(foreach dir,$(XPCSHELL_TESTS),$(testxpcobjdir)/$(MODULE)/$(dir))
>+	  $(testxpcsrcdir)/runxpcshelltests.py \
>+	  --symbols-path=$(DIST)/crashreporter-symbols \
>+	  --test-path=$(SOLO_FILE) \
>+	  --interactive \
>+	  $(DIST)/bin/xpcshell \
>+	  $(foreach dir,$(XPCSHELL_TESTS),$(testxpcobjdir)/$(MODULE)/$(dir))

Whitespace: TODO

> check-one:
...
>-          $(foreach dir,$(XPCSHELL_TESTS),$(testxpcobjdir)/$(MODULE)/$(dir))

Same
> ifndef LIBRARY
> ifdef STATIC_LIBRARY_NAME
>+ifneq (,$(filter OS2 WINNT WINCE,$(OS_ARCH)))
>+ifdef SHORT_LIBNAME
>+STATIC_LIBRARY_NAME	:= $(SHORT_LIBNAME)
>+SHARED_LIBRARY_NAME	:= $(SHORT_LIBNAME)
>+endif
>+endif
> LIBRARY			:= $(LIB_PREFIX)$(STATIC_LIBRARY_NAME).$(LIB_SUFFIX)
> endif # STATIC_LIBRARY_NAME
> endif # LIBRARY

removed from m-c in Bug 514665, Adjusted in OS2 Bug 520418, and moved away from here in Bug 536451.

Tracked by Bug 534408 (overloaded in scope slightly).
Blocks: 534408
No longer blocks: 534408
Depends on: 534408
I planned on doing this, but later when the ports would be (even) more complete:
I'm fine that you take over :->

Whitespaces: fwiw, usually I tried to improve it on c-c and Ted refused to sync' on m-c...
Flags: in-testsuite-
As I was checking that other bug...
Attachment #438670 - Flags: review?(bugspam.Callek)
Comment on attachment 438670 [details] [diff] [review]
(Av1) Fix whitespace nits from bug 526760

At least for our end, unless we are touching the lines already I'd rather do the whitespace all in one go, and match to m-c; even if m-c whitespace is flawed. [I would be OK with you doing this + the ones I already mentioned so far here, but would rather wait a bit]
Attachment #438670 - Flags: review?(bugspam.Callek) → review-
(In reply to comment #5)
> (From update of attachment 438670 [details] [diff] [review])
> At least for our end, unless we are touching the lines already I'd rather do
> the whitespace all in one go, and match to m-c; even if m-c whitespace is
> flawed.

Actually, in the latter case, I'd challenge you guys to actually do a patch for m-c to correct it and have both versions use non-flawed whitespace then. :)
newer m-c vs c-c diff....
Assignee: nobody → bugspam.Callek
Attachment #438233 - Attachment is obsolete: true
Status: NEW → ASSIGNED
actually output the diff to correct dir.
Attachment #464280 - Attachment is obsolete: true
Attached patch v1Splinter Review
This encompasses all syncs from the m-c version that are not intentional changes

The whitespace changes that are better on our side I left alone.

I also did not port over the fakelibs change in this patch, that will likely be best in its own bug. [given how many followup csets there were, and the potential likelihood of breakage]
Attachment #438670 - Attachment is obsolete: true
Attachment #464287 - Flags: review?(bugzilla)
Comment on attachment 464287 [details] [diff] [review]
v1

Whoever reviews first...
Attachment #464287 - Flags: review?(kairo)
Could you please make sure that all things it ports are marked on http://dev.seamonkey.at/?d=x&i=mozilla&m=c and hopefully also marked as "ported" when you check this in?
That way we can ensure things we still need to port are correctly tracked.
Comment on attachment 464287 [details] [diff] [review]
v1

Ftr, this patch would restore code(s) which was previously removed,
as in http://hg.mozilla.org/comm-central/diff/b90af1833692/config/rules.mk
for example :-/
You may want to check/confirm what the current situation is...
(In reply to comment #11)
> Could you please make sure that all things it ports are marked on
> http://dev.seamonkey.at/?d=x&i=mozilla&m=c and hopefully also marked as
> "ported" when you check this in?
> That way we can ensure things we still need to port are correctly tracked.

Planned, was to do that within the next two days.

(In reply to comment #12)
> Comment on attachment 464287 [details] [diff] [review]
> v1
> 
> Ftr, this patch would restore code(s) which was previously removed,
> as in http://hg.mozilla.org/comm-central/diff/b90af1833692/config/rules.mk
> for example :-/
> You may want to check/confirm what the current situation is...

Your cset link I think is the only "intentially removed" stuff I added back. I couldn't recall if or when it was removed, so I just added it back. Easy to take back out.
Comment on attachment 464287 [details] [diff] [review]
v1

r=Standard8 without the unintentionally added back of the intentionally removed stuff.

I think it would be useful to get the fakelibs stuff ported asap as well - I've seen weird failures etc on Windows over the last week or so and I think something dependency related is causing it (though I'm not sure what). So co-ordinating that landing would be good as well.
Attachment #464287 - Flags: review?(kairo)
Attachment #464287 - Flags: review?(bugzilla)
Attachment #464287 - Flags: review+
I took the liberty of a+ ing myself, given that this has little to NO change of breaking MozMill on the closed tree.

(In reply to comment #13)
> (In reply to comment #11)
> > Could you please make sure that all things it ports are marked on
> > http://dev.seamonkey.at/?d=x&i=mozilla&m=c and hopefully also marked as
> > "ported" when you check this in?
> > That way we can ensure things we still need to port are correctly tracked.
> 
> Planned, was to do that within the next two days.

It appears I can't do that actually (Some perm bit got unticked on my login). That said, EVERY change on http://dev.seamonkey.at/?d=x&i=mozilla&m=c&c=f&f.i=16 (for rules.mk) is now either ported [or ignorable from m-c backouts]
Attaching the intentional (remaining) diffs between m-c and c-c. and obsoleting the earlier tracking diff.
Attachment #464282 - Attachment is obsolete: true
(and attach the right diff file. this time without appending to old diff).
Attachment #468103 - Attachment is obsolete: true
http://hg.mozilla.org/comm-central/rev/370088848500
Status: ASSIGNED → RESOLVED
Closed: 14 years ago
Resolution: --- → FIXED
(In reply to comment #15)
> It appears I can't do that actually (Some perm bit got unticked on my login).

Hmm, found an interesting bug in my system. The settings should be corrected now, you can make those changes now.
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Created:
Updated:
Size: