Support MOZ_TRACK_MODULE_DEPS on Win32

RESOLVED FIXED in mozilla0.9.5

Status

SeaMonkey
Build Config
RESOLVED FIXED
17 years ago
14 years ago

People

(Reporter: Alec Flett, Assigned: Alec Flett)

Tracking

Trunk
mozilla0.9.5
x86
Windows 2000

Firefox Tracking Flags

(Not tracked)

Details

Attachments

(17 attachments, 2 obsolete attachments)

1.31 KB, patch
cls
: review+
Details | Diff | Splinter Review
8.08 KB, patch
cls
: review+
Details | Diff | Splinter Review
5.98 KB, patch
cls
: review+
Details | Diff | Splinter Review
13.62 KB, patch
cls
: review+
Details | Diff | Splinter Review
34.82 KB, patch
cls
: review+
Details | Diff | Splinter Review
23.62 KB, patch
cls
: review+
Details | Diff | Splinter Review
19.44 KB, patch
cls
: review+
Details | Diff | Splinter Review
28.53 KB, patch
cls
: review+
Details | Diff | Splinter Review
15.59 KB, patch
cls
: review+
Details | Diff | Splinter Review
1.52 KB, patch
cls
: review+
Details | Diff | Splinter Review
17.70 KB, patch
cls
: review+
Details | Diff | Splinter Review
17.42 KB, patch
cls
: review+
Details | Diff | Splinter Review
2.07 KB, patch
Brian Ryner (not reading)
: review+
Alec Flett
: superreview+
Details | Diff | Splinter Review
36.75 KB, patch
Alec Flett
: superreview+
Details | Diff | Splinter Review
931 bytes, patch
cls
: review+
Details | Diff | Splinter Review
100.95 KB, patch
cls
: review+
Details | Diff | Splinter Review
5.72 KB, patch
cls
: review+
Details | Diff | Splinter Review
(Assignee)

Description

17 years ago
Since my primary machine is now (gasp!) Win32, I decided to add support for
module dependancies so that I could work on this stuff on my primary machine.

I'm attaching patches for the general build system, and then attaching patches
for individual components. Nothing is affected unless you have
MOZ_TRACK_MODULE_DEPS in your environment.
(Assignee)

Comment 1

17 years ago
Created attachment 48301 [details] [diff] [review]
add support in config.mak
(Assignee)

Comment 2

17 years ago
looking for r/sr= on that. I had to SLIGHTLY change the INCS line for the
general case, because I PUBLIC was being used not only as the destination for
exported files, but also the prefix to the nspr include files, which was bad.
(Assignee)

Comment 3

17 years ago
Created attachment 48303 [details] [diff] [review]
update string and xpcom

Comment 4

17 years ago
r=cls
(Assignee)

Updated

17 years ago
Status: NEW → ASSIGNED
Target Milestone: --- → mozilla0.9.5
(Assignee)

Comment 5

17 years ago
Created attachment 48447 [details] [diff] [review]
patches for intl
(Assignee)

Comment 6

17 years ago
Created attachment 48449 [details] [diff] [review]
update for mail

Updated

17 years ago
Attachment #48301 - Flags: review+

Updated

17 years ago
Attachment #48303 - Flags: review+

Updated

17 years ago
Attachment #48447 - Flags: review+

Updated

17 years ago
Attachment #48449 - Flags: review+
(Assignee)

Comment 7

17 years ago
Created attachment 48883 [details] [diff] [review]
update for layout, content, gfx, gfx2, widget, and modules/libpr0n
(Assignee)

Comment 8

17 years ago
cls, any chance I can get reviewson the last 4 patches?
(Assignee)

Comment 9

17 years ago
Created attachment 48892 [details] [diff] [review]
changes for jpeg, sun-java, db, netwerk, etc

Comment 10

17 years ago
Comment on attachment 48883 [details] [diff] [review]
update for layout, content, gfx, gfx2, widget, and modules/libpr0n

r=cls
Attachment #48883 - Flags: review+

Comment 11

17 years ago
Comment on attachment 48892 [details] [diff] [review]
changes for jpeg, sun-java, db, netwerk, etc

r=cls
Attachment #48892 - Flags: review+
(Assignee)

Comment 12

17 years ago
ok, the above patches have landed... I'm about 90% done with this work, I'll get
the rest of the patches ready today.
(Assignee)

Comment 13

17 years ago
Created attachment 48983 [details] [diff] [review]
next set of diffs
(Assignee)

Comment 14

17 years ago
Created attachment 48984 [details] [diff] [review]
and yet another set.
(Assignee)

Comment 15

17 years ago
Created attachment 48985 [details] [diff] [review]
And finally, the xpfe directory
(Assignee)

Comment 16

17 years ago
Created attachment 48986 [details] [diff] [review]
slight changes I needed for the dependancy tool

Comment 17

17 years ago
Comment on attachment 48983 [details] [diff] [review]
next set of diffs

r=cls
Attachment #48983 - Flags: review+

Comment 18

17 years ago
Comment on attachment 48984 [details] [diff] [review]
and yet another set.

r=cls
Attachment #48984 - Flags: review+

Comment 19

17 years ago
Comment on attachment 48985 [details] [diff] [review]
And finally, the xpfe directory

r=cls
Attachment #48985 - Flags: review+

Comment 20

17 years ago
Comment on attachment 48986 [details] [diff] [review]
slight changes I needed for the dependancy tool

r=cls
Attachment #48986 - Flags: review+
(Assignee)

Comment 21

17 years ago
Created attachment 49098 [details] [diff] [review]
argh. additional updates to mailnews

Comment 22

17 years ago
Comment on attachment 49098 [details] [diff] [review]
argh. additional updates to mailnews

r=cls
Attachment #49098 - Flags: review+
(Assignee)

Comment 23

17 years ago
Created attachment 49144 [details] [diff] [review]
some cleanups - finding lots of stray LINCS=

Comment 24

17 years ago
Comment on attachment 49144 [details] [diff] [review]
some cleanups - finding lots of stray LINCS=

r=cls
Attachment #49144 - Flags: review+

Comment 25

17 years ago
Created attachment 49183 [details] [diff] [review]
More cleanup, mostly tests & renaming some mods (raptor? feh)

Comment 26

17 years ago
Created attachment 49184 [details] [diff] [review]
security/ portion of cleanup

Comment 27

17 years ago
Comment on attachment 49183 [details] [diff] [review]
More cleanup, mostly tests & renaming some mods (raptor? feh)

Umm, apparently, I missed somethings. Trying again.
Attachment #49183 - Attachment is obsolete: true

Comment 28

17 years ago
Created attachment 49189 [details] [diff] [review]
Updated tests, mod-renaming sans raptor patch
(Assignee)

Comment 29

17 years ago
argh... I just finished all this work in my local tree.. can we coordinate our work?
(Assignee)

Comment 30

17 years ago
Comment on attachment 49189 [details] [diff] [review]
Updated tests, mod-renaming sans raptor patch

sr=alecf
Attachment #49189 - Flags: superreview+
(Assignee)

Comment 31

17 years ago
actually, can you hold off on the raptor renaming? My tree also contains
seperation of layout vs. layout_xul, content vs. content_xul, and merging of
"xuldoc" and "xultmpl" into layout_xul
(Assignee)

Comment 32

17 years ago
Created attachment 49269 [details] [diff] [review]
massive cleanup of raptor/content/layout/etc, and LINCS
(Assignee)

Comment 33

17 years ago
those are the final changes that I need to land in order to mark this bug FIXED.

Changes include:
- moved everything in content/ over to "content" and content_xul - I
special-cased everything in content/xul to content_xul because it's the largest
chunk, and this will hopefully make it easy to seperate out. content_xul also
got the stuff from xuldoc and xultmpl
- did a similar xul-oriented cleanup in layout, creating layout and layout_xul
- I did a huge sweep through the tree with grep and friends, to clean up any
LINCS lines that directly reference the dist directory. There were a few odd
directories like under sun-java that I didn't touch, because they have their own
wierd rules.
- After cleaning up LINCS, I found many instances of MODULE= definitions that
came AFTER the inclusion of config.mak, so I moved that. I also found a few
stray items missing from REQUIRES

That's it. Reviews?

Comment 34

17 years ago
After removing all of my changes, pulling a fresh tree and applying your latest
patch, I'm dying in xpcom\typelib\xpt\tests becaues a file cannot find xpt_xdr.h
(Assignee)

Comment 35

17 years ago
how wierd! somehow my dependencies just won't make this directory build.
anyway, just add
REQURIES=xpcom

to that makefile.win
I've done so in my local tree..

I'm trying (another) clobber to see if there are any others that I've missed.
(Assignee)

Comment 36

17 years ago
Created attachment 49337 [details] [diff] [review]
better way of handling REQUIRES without perl
(Assignee)

Updated

17 years ago
Attachment #49269 - Attachment is obsolete: true
(Assignee)

Comment 37

17 years ago
Created attachment 49392 [details] [diff] [review]
take two on massive cleanup patch
(Assignee)

Comment 38

17 years ago
sure enough, my builds weren't building any .exe's - normally there's a rule
that rebuilds a directory if rules.mak or config.mak changes - doesn't happen
for .exes.

Anyway, that patch fixes everything - allows me to do a complete clobber build.
reviews?
(Assignee)

Comment 39

17 years ago
Comment on attachment 49184 [details] [diff] [review]
security/ portion of cleanup

almost forgot...sr=alecf
Attachment #49184 - Flags: superreview+

Comment 40

17 years ago
Comment on attachment 49392 [details] [diff] [review]
take two on massive cleanup patch

r=cls w/ the changes for intl\chardet\src\{windows,classic}\makefile.win
Attachment #49392 - Flags: review+

Comment 41

17 years ago
wtc or javi: can either of you review & checkin attachment 49184 [details] [diff] [review] ?  Thanks.

Comment 42

17 years ago
cls:  I reviewed the patch in attachment 49184 [details] [diff] [review].  The changes
to the two makefiles.win's under security/manager are fine.
You might want to ask someone to verify the modules listed
in REQUIRES are indeed required.  I am not qualified to verify
that.

I will check in the change for security/nss/makefile.win when
the tree re-opens.  (You have cvs commit access to security/manager.)

By the way, javi is on sabbatical now.
(Assignee)

Comment 43

17 years ago
the final patch has been checked in. 

cls: Can I get a review on attachment 49337 [details] [diff] [review]? it should speed up the windows
REQUIRES build.. then we'll be ready to turn it on!

Comment 44

17 years ago
I checked in the NSS portion of attachment 49184 [details] [diff] [review].
The PSM portion of that patch still needs to be
reviewed (for the REQUIRES lists) and checked in.

Comment 45

17 years ago
Comment on attachment 49337 [details] [diff] [review]
better way of handling REQUIRES without perl

Oops. Thought I already did this. r=cls
Attachment #49337 - Flags: review+

Comment 46

17 years ago
For review of security/manager parts of attachment 49184 [details] [diff] [review], please just ask
someone who has insight into the windows makefiles.

If you are sure it works, moa=kaie.
I can check the changes in if you want me to.
Comment on attachment 49184 [details] [diff] [review]
security/ portion of cleanup

r=bryner
Attachment #49184 - Flags: review+
(Assignee)

Comment 48

17 years ago
all the appropriate patches have now been checked in. marking fixed. yay!
Status: ASSIGNED → RESOLVED
Last Resolved: 17 years ago
Resolution: --- → FIXED

Comment 49

17 years ago
Have they? I'm still showing at least the dbm headers still being placed into
the main inclue dir and the non-perl version of config.mak hasn't been checked in.
(Assignee)

Comment 50

17 years ago
oof! you're right. I had timestamped my config.mak to last week to avoid
building - I didn't realize CVS would thus ignore the file!
the non-perl config.mak has been checked in.
I guess I assumed the dbm stuff was outdated (nobody uses dbm) so it didn't
matter where it went. I don't believe we should actually export those files at
all...
(Assignee)

Comment 51

17 years ago
Created attachment 50252 [details] [diff] [review]
additional patches from jonsmirl@mediaone.net

Comment 52

17 years ago
Comment on attachment 50252 [details] [diff] [review]
additional patches from jonsmirl@mediaone.net

r=cls
Attachment #50252 - Flags: review+

Comment 53

17 years ago
Well, at least NSS & necko (cache) still use dbm.  And as I just filed in bug
100966, xp_file.h depends upon dbm/winfile.h so there are a few more win32
dependencies that should be cleaned up.

Status: RESOLVED → REOPENED
Resolution: FIXED → ---
(Assignee)

Comment 54

17 years ago
jon, I've attached your svg patch here, but I tried applying it and found that
some of the svg directories were missing.. then looking at the patch, I saw that
they were against Attic files.. did you make that patch against a branch? or is
there a special way of checking out the svg stuff?
Status: REOPENED → ASSIGNED
(Assignee)

Comment 55

17 years ago
jon says:

The SVG code lives on a branch: SVG_20010721_BRANCH
I filed this bug directly against the SVG people:
http://bugzilla.mozilla.org/show_bug.cgi?id=101139
 
I wish this branch would get merged into the trunk it's getting to be a lot of
code on it. There is an internal Netscape person who maintains the branch; I
think his name is Bradley. The mozilla news server is down so I can't check.

Comment 56

17 years ago
Sorry, I thought I commented on attachment 50252 [details] [diff] [review]. I checked in a modified trunk
version this weekend.  bbaetz is no longer interning at Netscape.

 
(Assignee)

Comment 57

17 years ago
oh, thanks cls!
(Assignee)

Comment 58

17 years ago
anyone mind if I mark this fixed now? config.mak has been checked in, the builds
are going...
Status: ASSIGNED → RESOLVED
Last Resolved: 17 years ago17 years ago
Resolution: --- → FIXED
Product: Browser → Seamonkey
You need to log in before you can comment on or make changes to this bug.