754358 - per-directory xpt linking race condition?

Status: RESOLVED WORKSFORME

(Firefox Build System :: General, defect)

Description

[(not currently active) Ted Mielczarek]

I've seen two occurances of an xpt.py error that looks like some kind of race condition:

Traceback (most recent call last):
  File "/scratch/work/builds/mozilla-central/mozilla-central.10-12-09.12-05/obj/dist/sdk/bin/xpt.py", line 1382, in <module>
    xpt_link(sys.argv[3:]).write(sys.argv[2])
  File "/scratch/work/builds/mozilla-central/mozilla-central.10-12-09.12-05/obj/dist/sdk/bin/xpt.py", line 1276, in xpt_link
    t = read_input(f)
  File "/scratch/work/builds/mozilla-central/mozilla-central.10-12-09.12-05/obj/dist/sdk/bin/xpt.py", line 1264, in read_input
    return Typelib.read(i)
  File "/scratch/work/builds/mozilla-central/mozilla-central.10-12-09.12-05/obj/dist/sdk/bin/xpt.py", line 1083, in read
    data_pool_offset) = Typelib._header.unpack(data[:Typelib._header.size])
struct.error: unpack requires a string argument of length 32
make[7]: *** [_xpidlgen/necko.xpt] Error 1
make[6]: *** [libs] Error 2
make[5]: *** [libs] Error 2
make[4]: *** [libs_tier_platform] Error 2
make[3]: *** [tier_platform] Error 2
make[2]: *** [default] Error 2
make[1]: *** [realbuild] Err

xpt.py is erroring because it's trying to read an xpt file that apparently hasn't been written yet (or hasn't been fully written out?) The build system attempts to link together all the XPT files in a directory:
http://mxr.mozilla.org/mozilla-central/source/config/rules.mk#1444

The dependencies there all look correct, so I'm not sure what's going on.

Comment 1

[Daniel Holbert [:dholbert] (vacation until Jun 23)]

I hit this error twice today -- I did a 'hg pull -u', kicked off an incremental build, [error], did some unrelated qpopping & qpushing, then built again, [error].

Then I clobbered & was able to build successfully.

Comment 2

[(not currently active) Ted Mielczarek]

Joey: I feel like you filed a bug about this but I can't remember what it was offhand.

Comment 3

[Joey Armstrong [:joey]]

(In reply to Ted Mielczarek [:ted.mielczarek] from comment #2)
> Joey: I feel like you filed a bug about this but I can't remember what it
> was offhand.

The other race condition bug was 809978

Comment 4

[Gregory Szorc [:gps]]

Nobody has complained about this in a few years. I'm going to say this no longer exists. If it does, chances are the XPIDL refactoring I'm doing will squash it.