619006 - [SeaMonkey] 'Linux comm-* build' (silently) reports "Error: 'objdir/dist/bin/libxul.so': No such file", since upgrade to builbot 0.8

Status: VERIFIED FIXED

(SeaMonkey :: Release Engineering, defect, P1)

Description

[Serge Gautherie (:sgautherie)]

http://tinderbox.mozilla.org/showlog.cgi?log=SeaMonkey2.0/1292221103.1292225142.6235.gz
Linux comm-1.9.1 build on 2010/12/12 22:18:23
{
======== BuildStep started ========
no change
=== Output ===
python /builds/slave/comm-1.9.1-linux/tools/buildfarm/utils/count_ctors.py objdir/dist/bin/libxul.so
...
readelf: Error: 'objdir/dist/bin/libxul.so': No such file
None	objdir/dist/bin/libxul.so
program finished with exit code 0
}

Related to builbot change to v0.8 !?

Comment 1

[Serge Gautherie (:sgautherie)]

http://tinderbox.mozilla.org/showlog.cgi?log=SeaMonkey/1292285113.1292286316.6008.gz
Linux comm-central-trunk build on 2010/12/13 16:05:13

Trunk too.

Comment 2

[Serge Gautherie (:sgautherie)]

Hum,

http://tinderbox.mozilla.org/showlog.cgi?log=SeaMonkey/1292287877.1292294137.10095.gz
Linux comm-central-trunk build on 2010/12/13 16:51:17
has the same failure, yet is green :-/

http://tinderbox.mozilla.org/showlog.cgi?log=SeaMonkey/1292283352.1292287124.9342.gz
Linux x86-64 comm-central-trunk build on 2010/12/13 15:35:52
too.

MacOSX and Windows seem fine.

Comment 3

[Serge Gautherie (:sgautherie)]

What makes builds actually "fail" is in fact
Bug 619009 - Lots of "graph server post results complete failed", since 2010.12.12

Comment 4

[Serge Gautherie (:sgautherie)]

(In reply to comment #0)
> python /builds/slave/comm-1.9.1-linux/tools/buildfarm/utils/count_ctors.py
> objdir/dist/bin/libxul.so

Ah, I guess this should rather be objdir/mozilla/dist/bin/libxul.so.

See
http://mxr.mozilla.org/build/search?string=libxul.so&case=1&find=%2Fprocess%2Ffactory.py

Comment 5

[Serge Gautherie (:sgautherie)]

Attachment: (Av1) s/objdir/mozillaObjdir/ [Duplicated by bug 619975]

Untested, but copied from a few lines below...

Comment 6

[Justin Wood (:Callek)]

Comment on attachment 497437 [details] [diff] [review]
(Av1) s/objdir/mozillaObjdir/
[Duplicated by bug 619975]

I think we want KaiRo to peek at this too...

Comment 7

[Serge Gautherie (:sgautherie)]

(In reply to comment #2)
> same failure, yet is green :-/

Ben, shouldn't this step have warnOnFailure=True and/or haltOnFailure=True or the like too? So that this error could be explicitly noticed...

Comment 8

[Robert Kaiser]

Comment on attachment 497437 [details] [diff] [review]
(Av1) s/objdir/mozillaObjdir/
[Duplicated by bug 619975]

That should be it, yes, as it's a factory used by everyone, Ben needs to do the review and we'll need to do a coordinated checkin, though.

Comment 9

[Serge Gautherie (:sgautherie)]

Nightlies are affected too:
http://tinderbox.mozilla.org/showlog.cgi?log=SeaMonkey2.0/1292140801.1292144742.23936.gz&fulltext=1
Linux x86-64 comm-1.9.1 nightly on 2010/12/12 00:00:01
s: cb-seamonkey-linux64-01
{
readelf: Error: 'objdir/dist/bin/libxul.so': No such file
}

Comment 10

[[e-mail] (:bhearsum)]

Comment on attachment 497437 [details] [diff] [review]
(Av1) s/objdir/mozillaObjdir/
[Duplicated by bug 619975]

Looks ok, needs testing in staging.

Comment 11

[Serge Gautherie (:sgautherie)]

checkin-needed, or should I simply push it to default?

***

Ben, what about comment 7?

Comment 12

[[e-mail] (:bhearsum)]

(In reply to comment #11)
> checkin-needed, or should I simply push it to default?
> 
> ***
> 
> Ben, what about comment 7?

No, please don't land it anywhere or mark checkin-needed yet. I'll test this in staging next time I'm doing something else there.

Comment 13

[Serge Gautherie (:sgautherie)]

(In reply to comment #12)
> I'll test this in staging next time I'm doing something else there.

Ping.

Comment 14

[Serge Gautherie (:sgautherie)]

Yet another case of stalled process which leads to double work (in bug 619975) :-(

Comment 15

[Justin Wood (:Callek)]

based on c#14 and that this bug wasn't closed (yet) I'm now morphing it into a SeaMonkey releng bug to track the deployment of this buildbot update. Will resolve when its resolved on our production.

Comment 16

[Robert Kaiser]

I just updated buildbotcustom as I checked in the last patch to the repo that we had applied locally, and with the reconfig I also did afterwards, this should be fixed.

Comment 17

[Serge Gautherie (:sgautherie)]

On
http://tinderbox.mozilla.org/showbuilds.cgi?tree=SeaMonkey&maxdate=1294810489&hours=24&legend=0&norules=1

http://tinderbox.mozilla.org/showlog.cgi?log=SeaMonkey/1294758558.1294764156.15351.gz
Linux comm-central-trunk build on 2011/01/11 07:09:18
{
readelf: Error: 'objdir/dist/bin/libxul.so': No such file
...
TinderboxPrint: num_ctors: unknown
}

http://tinderbox.mozilla.org/showlog.cgi?log=SeaMonkey/1294767933.1294776098.5864.gz
Linux comm-central-trunk build on 2011/01/11 09:45:33
{
329	objdir/mozilla/dist/bin/libxul.so
...
TinderboxPrint: num_ctors: 329
}

V.Fixed (on trunk)

Comment 18

[Serge Gautherie (:sgautherie)]

Yet, this is still not fixed on SeaMonkey 2.0...

Comment 19

[Serge Gautherie (:sgautherie)]

(In reply to comment #7)
> Ben, shouldn't this step have warnOnFailure=True and/or haltOnFailure=True or
> the like too? So that this error could be explicitly noticed...

I filed bug 626232.