Bug 15099 - JvCreateJavaVM(NULL) make string to float conversion fail
Summary: JvCreateJavaVM(NULL) make string to float conversion fail
Status: RESOLVED WORKSFORME
Alias: None
Product: gcc
Classification: Unclassified
Component: libgcj (show other bugs)
Version: 3.3.1
: P2 normal
Target Milestone: ---
Assignee: Not yet assigned to anyone
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2004-04-23 12:08 UTC by Johan Baltié
Modified: 2005-07-23 22:49 UTC (History)
2 users (show)

See Also:
Host: i586-mandrake-linux
Target:
Build:
Known to work:
Known to fail:
Last reconfirmed:

Attachments
Preprocessed file. (102.35 KB, text/plain)
2004-04-23 12:10 UTC, Johan Baltié 		Details
Program that does create the bug (242 bytes, text/plain)
2004-04-23 12:29 UTC, Johan Baltié 		Details
View All Add an attachment (proposed patch, testcase, etc.)

Note You need to log in before you can comment on or make changes to this bug.
Description Johan Baltié 2004-04-23 12:08:50 UTC 
The program:

int main()
{
    string test1 = "3.859155991718317e-06";


    cout << test1
         << " - "
         << test1.c_str()
         << " - "
         << std::atof(test1.c_str()) << endl;
    JvCreateJavaVM(__null);
    cout << test1
         << " - "
         << test1.c_str()
         << " - "
         << std::atof(test1.c_str()) << endl;
    JvDetachCurrentThread();

    cout << test1
         << " - "
         << test1.c_str()
         << " - "
         << std::atof(test1.c_str()) << endl;


  return (0);
}

gives as an output:

3.859155991718317e-06 - 3.859155991718317e-06 - 3.85916e-06
3.859155991718317e-06 - 3.859155991718317e-06 - 3
3.859155991718317e-06 - 3.859155991718317e-06 - 3

Clearly once the VM is created no string to float conversion works.
Comment 1 Johan Baltié 2004-04-23 12:10:47 UTC 
Created attachment 6150 [details]
Preprocessed file.

Result of "g++ -save-temps -I /usr/include/libgcj-3.3.1 -lgcj test.cc"
Comment 2 Andrew Pinski 2004-04-23 12:26:44 UTC 
Can you provide the __non__ preprocessed source?  Also can you try 3.4.0 while you are at it?
Comment 3 Johan Baltié 2004-04-23 12:29:03 UTC 
Created attachment 6151 [details]
Program that does create the bug

Output of this program:

3.859155991718317e-06 - 3.859155991718317e-06 - 3.85916e-06
3.859155991718317e-06 - 3.859155991718317e-06 - 3
3.859155991718317e-06 - 3.859155991718317e-06 - 3
Comment 4 Johan Baltié 2004-04-23 12:33:37 UTC 
I cannot use the 3.4.0 version for now because I'm not "able" to install it in a
proper way (I'm not admin and installing a new compiler in a local already led
to problem last time we've done it).
Comment 5 Andrew Pinski 2004-04-23 12:48:17 UTC 
I cannot reproduce it on i686-pc-linux-gnu running Redhat 7.3, running glibc 2.2.5 on 3.4.0, 3.3.3 or 
3.5.0, so this is problem in your glibc because atof should be using the C locale always (read the C 
standard). There is no way around it, sorry except for updating your version of glibc.