Bug 145886 - Add support for Linux (PPC/GTK 2) - liblocalfile_1_0_0.so
Summary: Add support for Linux (PPC/GTK 2) - liblocalfile_1_0_0.so
Status: RESOLVED FIXED
Alias: None
Product: Platform
Classification: Eclipse Project
Component: Resources (show other bugs)
Version: 3.2   Edit
Hardware: Power PC Linux-GTK
: P3 normal (vote)
Target Milestone: 3.3 M1   Edit
Assignee: Platform-Resources-Inbox CLA
QA Contact:
URL:
Whiteboard:
Keywords: contributed
: 137845 (view as bug list)
Depends on: 146080
Blocks:
  Show dependency tree
 
Reported: 2006-06-07 18:29 EDT by Corey Ashford CLA
Modified: 2006-08-24 12:15 EDT (History)
2 users (show)

See Also:


Attachments
liblocal.so file for Linux-ppc (SLES 10 RC2) (38.95 KB, application/octet-stream)
2006-06-08 14:18 EDT, Corey Ashford CLA
no flags Details
liblocalfile built on SLES 9 instead of SLES 10 (40.09 KB, application/octet-stream)
2006-08-22 18:10 EDT, Corey Ashford CLA
no flags Details

Note You need to log in before you can comment on or make changes to this bug.
Description Corey Ashford CLA 2006-06-07 18:29:32 EDT
Recently, I've been running the Eclipse JUnit Plugin Tests on Eclipse 3.2RC3 and 3.2RC7 for Linux (SLES 10 RC2) on PowerPC and have run into the issue that some tests were failing due to a missing liblocalfile_1_0_0.so file.  Once I built this library, and put it on my LD_LIBRARY_PATH, the tests passed.

A pre-compiled .so is available for other OS/architecture combinations from CVS, but is not for linux-gtk-ppc.

What I think is needed is to arrange to have this library built in the build process and committed into CVS.  I don't know if there is more mechanism that needs to change to get the library into the Eclipse nightly (etc.) build distributions.
Comment 1 John Arthorne CLA 2006-06-08 13:13:16 EDT
Would you mind attaching the recompiled library to this bug?  Were any changes required to the source (makefile, C file, header file)?
Comment 2 Corey Ashford CLA 2006-06-08 14:18:01 EDT
Created attachment 43901 [details]
liblocal.so file for Linux-ppc (SLES 10 RC2)

I have attached the liblocalfile_1_0_0.so file I built for SLES 10 RC2 for PowerPC.  No modifications to the source or Makefile were needed.  I just invoked the make command and it built without a complaint.
Comment 3 John Arthorne CLA 2006-06-08 14:35:06 EDT
Thanks Corey.  I have put in a request to the webmaster to have a project created in the repository for this fragment. Just to confirm, what version/build was this compiled library created from? Latest in HEAD?  The native source hasn't changed since October 2005, so if it was built from sources after that date it should be fine.
Comment 4 Corey Ashford CLA 2006-06-08 14:52:05 EDT
Yes, the source was all from the latest in HEAD as of a few days ago.

Thanks.
Comment 5 John Arthorne CLA 2006-06-14 16:32:35 EDT
A new entry has been added to core.map:

fragment@org.eclipse.core.filesystem.linux.ppc=v20060614,:pserver:anonymous@dev.eclipse.org:/cvsroot/eclipse,
Comment 6 Kim Moir CLA 2006-06-14 17:00:31 EDT
added to platform feature and running test build.
Comment 7 Kim Moir CLA 2006-06-14 18:00:23 EDT
test build was okay
Comment 8 John Arthorne CLA 2006-06-15 09:28:44 EDT
Thanks Kim for the test build, and thanks Corey for the library contribtion.  Marking fixed.
Comment 9 John Arthorne CLA 2006-07-24 14:36:39 EDT
*** Bug 137845 has been marked as a duplicate of this bug. ***
Comment 10 Corey Ashford CLA 2006-08-22 18:10:35 EDT
Created attachment 48422 [details]
liblocalfile built on SLES 9 instead of SLES 10

There were some reports of the library I had posted to this bug as not working right on some platforms.  Speculating that it might be because I built it on SLES 10, I have rebuilt it on SLES 9 in hopes that this will fix any compatibility problems.
Comment 11 John Arthorne CLA 2006-08-23 12:54:39 EDT
Thanks Corey, I have submitted this change to HEAD and we'll see if it solves the problems. SLES 9 is our target platform in the 3.3 plan, so that's what we should be compiling against.
Comment 12 John Arthorne CLA 2006-08-24 12:15:05 EDT
Just confirming that the latest patch does indeed fix the errors we were having. Thanks Corey.