Issue 113632 - DEV300_m86, unxlngx: getopt -s unknown option - breaks the build with --html
Summary: DEV300_m86, unxlngx: getopt -s unknown option - breaks the build with --html
Status: CLOSED FIXED
Alias: None
Product: Build Tools
Classification: Code
Component: code (show other issues)
Version: DEV300m84
Hardware: All All
: P1 (highest) Trivial (vote)
Target Milestone: ---
Assignee: hjs
QA Contact: issues@tools
URL:
Keywords: oooqa
: 113773 (view as issue list)
Depends on:
Blocks:
 
Reported: 2010-08-04 06:45 UTC by pavel
Modified: 2010-10-17 16:59 UTC (History)
6 users (show)

See Also:
Issue Type: PATCH
Latest Confirmation in: ---
Developer Difficulty: ---

Attachments
test if LIBRARY_PATH is not empty before using it (920 bytes, patch)
2010-08-06 06:34 UTC, dtardon 		no flags Details | Diff
updated patch (970 bytes, patch)
2010-08-06 10:59 UTC, dtardon 		no flags Details | Diff
Add an attachment (proposed patch, testcase, etc.)

Note You need to log in before you can comment on or make changes to this issue.
Description pavel 2010-08-04 06:45:24 UTC 
Hi,

DEV300_m86 on unxlngx in libegg and vcl:

Making:    libeggtraylx.so
getopt: invalid option -- s
Checking DLL ../unxlngx6.pro/lib/check_libeggtraylx.so ...: ok

The problem comes from this probably:

.IF "$(OS)$(CPU)" == "LINUXX"
EXTRALIBPATHS+=-L$(LIBRARY_PATH)
.ENDIF # "$(OS)$(CPU)" == "LINUXX"
Comment 1 dtardon 2010-08-06 06:33:23 UTC 
The LIBRARY_PATH variable is empty on my system (x86_64, Fedora Rawhide), so
there is "-L -- -s" on checkdll.sh's command line, which causes the error message.
Comment 2 dtardon 2010-08-06 06:34:51 UTC 
Created attachment 70980 [details]
test if LIBRARY_PATH is not empty before using it
Comment 3 vg 2010-08-06 09:51:15 UTC 
@dtargon: I would prefer:
.IF "$(OS)$(CPU)" == "LINUXX" && "$(LIBRARY_PATH)" != ""
condition
Comment 4 dtardon 2010-08-06 10:59:48 UTC 
Created attachment 70985 [details]
updated patch
Comment 5 pavel 2010-08-11 16:14:10 UTC 
*** Issue 113773 has been marked as a duplicate of this issue. ***
Comment 6 lohmaier 2010-08-29 00:54:35 UTC 
PING

m87 still with the bug :-(

I now get it as well on the buildbot, as apparently build.pl was modified or 
some modification in the botmaster causes it, don't know.

build alone doesn't trigger it (it prints the error, but doesn't fail to build, 
exit status of build is 0), but when using
build --html [--dontgraboutput] it fails

That difference in behaviour alone is worth a P1
Comment 7 lohmaier 2010-08-29 01:05:59 UTC 
Change that makes it fail now is the "use verbose in html mode", thus

build -- verbose=true  (or if you prefer VERBOSE=1 build)

will also trigger the build breaker (exit status 1)

http://pumbaa.ooodev.org:59145/source/diff/solenv/bin/build.pl?
r2=/solenv/bin/build.pl%40271452:cdf2145997d7&r1=/solenv/bin/build.pl%40271450:2
5c50c0a524e

(vgbugs10)
Comment 8 vg 2010-08-30 11:44:13 UTC 
fix will be in the next milestone
Comment 9 pavel 2010-09-21 20:26:51 UTC 
seen in m88...
Comment 10 Marcus 2010-10-17 16:59:04 UTC 
set to fixed
Comment 11 Marcus 2010-10-17 16:59:29 UTC 
closed