Bug 1979

Summary: size of unsigned int is wrong in ix86-common site file.
Product: [Build System, Metadata & Runtime] OE-Core Reporter: Joshua Immanuel <josh>
Component: coreAssignee: Nitin Kamble <nitin.a.kamble>
Status: RESOLVED FIXED QA Contact:
Severity: normal    
Priority: Medium CC: josh, meta.mr.watcher, meta.watcher, richard.purdie
Version: 1.1.1   
Target Milestone: 1.1.2   
Hardware: x86   
OS: Multiple   
Whiteboard:
OS type for building Yocto: --- Type of Regression: ---
Verified: Documentation change: ---
Attachments:
Description Flags
patch fixing the issue none

Description Joshua Immanuel 2012-02-10 04:38:17 UTC 
The size of unsigned int is defined as 1 byte in ix86-common site file. This causes the guile-native package to fail while checking for the unsigned int size in configure script.

I've already reported this in the mailing list at 

https://lists.yoctoproject.org/pipermail/yocto/2012-January/006707.html

For sake of keeping track. I am filing this bug.

I've attached the patch which solves the issue.
Comment 1 Joshua Immanuel 2012-02-10 04:40:10 UTC 
Created attachment 331 [details]
patch fixing the issue
Comment 2 Nitin Kamble 2012-02-10 14:54:32 UTC 
Following commit has fixed the issue in the master. which branch are you on?
Nitin


commit 80038a035241b9a1da32d4448fa34afcb697bad6
Author: Nitin A Kamble <nitin.a.kamble@intel.com>
Date:   Wed Jan 18 10:52:15 2012 -0800

    site/ix86-common: fix an error
    
    Fixed this line
    ac_cv_sizeof_unsigned_char=${ac_cv_sizeof_unsigned_int=1}
    
    as this line
    ac_cv_sizeof_unsigned_char=${ac_cv_sizeof_unsigned_char=1}
    
    This issue was causing guile recipe to compile-fail for x86 target.
    
    (From OE-Core rev: d71df3cc2ff2504d61078c578c0e73bbf53b6651)
    
    Signed-off-by: Nitin A Kamble <nitin.a.kamble@intel.com>
    Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

diff --git a/meta/site/ix86-common b/meta/site/ix86-common
index 512c9a0..e2796eb 100644
--- a/meta/site/ix86-common
+++ b/meta/site/ix86-common
@@ -1,6 +1,6 @@
 # general
 ac_cv_sizeof_char=${ac_cv_sizeof_char=1}
-ac_cv_sizeof_unsigned_char=${ac_cv_sizeof_unsigned_int=1}
+ac_cv_sizeof_unsigned_char=${ac_cv_sizeof_unsigned_char=1}
 ac_cv_sizeof_char_p=${ac_cv_sizeof_char_p=4}
 ac_cv_sizeof_unsigned_char_p=${ac_cv_sizeof_unsigned_char_p=4}
 ac_cv_sizeof_int=${ac_cv_sizeof_int=4}
Comment 3 Joshua Immanuel 2012-02-12 21:17:59 UTC 
Sorry for the delay in responding

I am in edison branch and my current head is adcf8bf7b52460b94998438e8c2bf854cdec0a80
Comment 4 Joshua Immanuel 2012-02-12 22:36:12 UTC 
Pulling from master into edison branch does not yield a clean merge. Many merge conflicts occur :(

So, cloning a master branch separately.
Comment 5 Nitin Kamble 2012-02-13 08:57:32 UTC 
Joshua,
If you are on edison branch, you can port the commit to edison, and send a commit pull request. "git cherry-pick " is your friend for this.

Nitin
Comment 6 Joshua Immanuel 2012-02-13 21:37:47 UTC 
Hello Nitin,

Thanks for the info. Will do a cherry-pick into my cloned edison branch.

I came across few more bugs in edison branch but those are irrelevant to the master branch (due to change of versions in the recipe). Should I file a separate bug report for the record sake?

Joshua
Comment 7 Nitin Kamble 2012-02-13 21:56:44 UTC 
Joshua,
  If those bugs bothers you, and you would like them to be fixed instead of going to master branch, then please file bug reports for them.

Thanks,
Nitin
Comment 8 Joshua Immanuel 2012-02-13 22:00:01 UTC 
Hello Nitin,

Ok. I'll file the bugs with relevant patches for them today.

Regards
Joshua
Comment 9 Joshua Lock - Disabled 2012-02-22 14:42:36 UTC 
Moving this to target 1.1.2, we should aim to merge the fix into the Edison branch ASAP.