204255 - ipcd builds tests even when --disable-tests is given

Status: RESOLVED FIXED

(Core :: IPC, defect, P3)

Description

[Conrad Carlen (not reading bugmail)]

Something like this:

ifdef ENABLE_TESTS
DIRS += tests
endif

needs to be added to ipc/ipcd/Makefile.in

Comment 1

[benc]

(sigh)

Comment 2

[Darin Fisher]

Attachment: v1 patch

this patch fixes/cleans-up several things:

1- honor ENABLE_TESTS build var
2- don't export intermediate static libs (bug 205025)
3- tmITransactionService should be ipcITransactionService
4- transaction service should not be a separate xpcom component library

plus some minor changes:

5- separate interfaces out into individual IDL file
6- get rid of *CID.h files (move CID/ContractID into IDL).. i think *CID.h
files are just overkill in this case.  there should either be just one ipcCID.h
file for all IDs or none.

Comment 3

[Conrad Carlen (not reading bugmail)]

Comment on attachment 122978 [details] [diff] [review]
v1 patch

>Index: ipc/ipcd/client/src/Makefile.in
>===================================================================
> 
> SHARED_LIBRARY_LIBS = \
>-			$(DIST)/lib/$(LIB_PREFIX)ipcdshared_s.$(LIB_SUFFIX) \
>-			$(DIST)/lib/$(LIB_PREFIX)ipcdlock_s.$(LIB_SUFFIX) \
>+			../../shared/src/$(LIB_PREFIX)ipcdshared_s.$(LIB_SUFFIX) \
>+			../../extensions/lock/src/$(LIB_PREFIX)ipcdlock_s.$(LIB_SUFFIX) \
>+			../../extensions/transmngr/src/$(LIB_PREFIX)transmngr_s.$(LIB_SUFFIX) \
>+			../../extensions/transmngr/common/$(LIB_PREFIX)transmngrcom_s.$(LIB_SUFFIX) \
> 			$(NULL)

Despite what I thought before, even if these static libs don't use
EXPORT_LIBRARY, they still end up in $(DIST)/lib so you don't need to use
relative paths. Fix that and r=ccarlen.


There's now another thing which will need change in mozilla: the manifests in
embedding/config. I had to, just today, include the ipc deamon and dlls (of
which there will be 1 less) in the manifests for windows and mach-o.

Comment 4

[Darin Fisher]

thanks conrad.. i'll fix up that Makefile as well as the manifests.

Comment 5

[Darin Fisher]

Attachment: v1.1 patch

Comment 6

[Darin Fisher]

Comment on attachment 123481 [details] [diff] [review]
v1.1 patch

carrying forward r=ccarlen.

alecf: can you please sr= these changes.  please see comment #2 for a summary
of these changes.  the patch is only large because i'm moving interfaces into
individual IDL files.  should be a quick review... thanks!

Comment 7

[Alec Flett]

Comment on attachment 123481 [details] [diff] [review]
v1.1 patch

sr=alecf

Comment 8

[Darin Fisher]

Comment on attachment 123481 [details] [diff] [review]
v1.1 patch

requesting approval for 1.4 final.  this change is mostly organizational, but i
would like very much to have these changes in place before we branch.  thx!

Comment 9

[(not reading, please use [e-mail] instead)]

Comment on attachment 123481 [details] [diff] [review]
v1.1 patch

a=sspitzer, assuming there are no packaging issues (for the trunk or embedding)
that are going to bite us.

Comment 10

[Darin Fisher]

fixed