Closed Bug 7106 Opened 26 years ago Closed 23 years ago

[PP] tinderbox doesn't report "breakage" when tests fail to build

Categories

(Webtools Graveyard :: Tinderbox, defect, P2)

x86
Linux
defect

Tracking

(Not tracked)

RESOLVED WONTFIX

People

(Reporter: sspitzer, Assigned: leaf)

Details

Look at http://cvs-mirror.mozilla.org/webtools/tinderbox/showlog.cgi?tree=SeaMonkey&errorparser=unix&buildname=shrike%20Linux/i386%202.2.5%20Clobber%20apprunner&buildtime=927731725&logfile=927731725.9502.gz this should be in flames, but it's not. here are the important bits from the end of the build log: ../../../../dist/include/jstypes.h:275: warning: ANSI C++ does not support `long long' ../../../../dist/include/jstypes.h:276: warning: ANSI C++ does not support `long long' c++ -o testimap ./imapProtocolTest.o -Wall -Wconversion -Wshadow -Wpointer-arith -Wbad-function-cast -Wcast-align -pedantic -pthread -g -L../../../../dist/./bin -L../../../../dist/./lib -L../../../../dist/./lib -L../../../../dist/./bin -labouturl -lfileurl -lftpurl -lgophurl -lremoturl -lhttpurl -lsockstuburl -lnetcache -lmimetype -lnetcnvts -lnetwork -lnetlib -lreg -lxpcom -lpwcac -lmozdbm -lxp -lpref -lmozjs -lraptorbase -lgmbasegtk -lsecfree -ljsdom -ljsurl -lraptorhtml -lraptorhtmlpars -lexpat -lxmltok -lraptorgfx -limg -lmozutil -lmsgbaseutil -ljpeg -lpng -lz -lplds3 -lplc3 -lnspr3 -lpthread -L/usr/lib -L/usr/X11R6/lib -lgtk -lgdk -rdynamic -lgmodule -lglib -ldl -lXext -lX11 -lm -lfl -lelf -lnsl -lutil -lresolv -ldl -lm /usr/bin/ld: cannot open -lraptorbase: No such file or directory collect2: ld returned 1 exit status NEXT gmake[5]: *** [testimap] Error 1 gmake[5]: Leaving directory `/builds/tinderbox/SeaMonkey/Linux_2.2.5_clobber/mozilla/mailnews/imap/tests/harness' NEXT gmake[4]: *** [install] Error 2 gmake[4]: Leaving directory `/builds/tinderbox/SeaMonkey/Linux_2.2.5_clobber/mozilla/mailnews/imap/tests' NEXT gmake[3]: *** [install] Error 2 gmake[3]: Leaving directory `/builds/tinderbox/SeaMonkey/Linux_2.2.5_clobber/mozilla/mailnews/imap' NEXT gmake[2]: *** [install] Error 2 gmake[2]: Leaving directory `/builds/tinderbox/SeaMonkey/Linux_2.2.5_clobber/mozilla/mailnews' NEXT gmake[1]: *** [install] Error 2 gmake[1]: Leaving directory `/builds/tinderbox/SeaMonkey/Linux_2.2.5_clobber/mozilla' NEXT gmake: *** [build] Error 2 /builds/tinderbox/SeaMonkey/Linux_2.2.5_clobber/./mozilla/dist/bin/apprunner exists, is nonzero, and executable. export binary exists, build successful. Testing... /builds/tinderbox/SeaMonkey/Linux_2.2.5_clobber/./mozilla/dist/bin/apprunner Success! /builds/tinderbox/SeaMonkey/Linux_2.2.5_clobber/./mozilla/dist/bin/apprunner is still running. Killing it, and its children,and its dog... ----------- Output from apprunner ---------------
Priority: P3 → P2
Target Milestone: M7
Summary: tinderbox doesn't report "breakage" when tests fail to build → [PP]tinderbox doesn't report "breakage" when tests fail to build
Target Milestone: M7 → M10
Steve says to move this to M10 for now. Can we find someone on the net to fix this?
Summary: [PP]tinderbox doesn't report "breakage" when tests fail to build → [PP] tinderbox doesn't report "breakage" when tests fail to build
Status: NEW → RESOLVED
Closed: 26 years ago
Resolution: --- → WORKSFORME
Hmm. Not sure what happened, but haven't seen the same problem since. Marking worksforme.
have the tests broken and its turned red? is apprunner the last thing to be built? Here's what I'm thinking: we build apprunner we continue on to build the tests, and it fails the script checks if the apprunner binary exists, if yes, we go on. we should also be checking the result of the gmake, to see if it succeeded, too.
Status: RESOLVED → REOPENED
Ok, I get it. This is a problem with the tinderbox build script, not tinderbox. Leaf, over to you.
Assignee: slamm → leaf
Status: REOPENED → NEW
Resolution: WORKSFORME → ---
Status: NEW → RESOLVED
Closed: 26 years ago26 years ago
Resolution: --- → LATER
Target Milestone: M10
Here is the crux of the problem. We used to build a monolithic binary. To test for success was very simple, looking for the executable. gmake error codes are not always bubbled to the top-level gmake, so checking for error there is not consistent. Since we live in a componentized world, it's very hard to check for each of them. We might be able to hook up some kind of manifest file checking thing based on the xpinstall stuff, but i'm not sure that is going to help us on unix. I'm going to unset the target milestone and later this, because i can't think of anything that doesn't involve a lot of manual work cataloguing what components need to exist in order for the app to be considered `fully built.'
LATER is deprecated per bug 35839.
Status: RESOLVED → REOPENED
Resolution: LATER → ---
resolved/wontfix.
Status: REOPENED → RESOLVED
Closed: 26 years ago23 years ago
Resolution: --- → WONTFIX
Product: Webtools → Webtools Graveyard
You need to log in before you can comment on or make changes to this bug.