Re: Bug with JNI on Solaris (wrong function address?)

[Rodrigo Augusto Barbato Ferreira <rodrigo dot ferreira at jewelvm dot com>]

Hi,
try specifying the nOptions field in JavaVMInitArgs record:

  vm_args.nOptions = 0;

since its not clear for me if JNI_GetDefaultJavaVMInitArgs
initializes this field (AFAIK Sun's implementation does not),
JNI_CreateJavaVM may be producing side effects.

Regards,
Rodrigo.

Martin Kahlert wrote:
> 
> Hi!
> I am back from vacation again and now i will continue with producing errors ;-)
> 
> This is the problem:
> $ cat bug.c
> 
> #include <jni.h>
> #include <stdio.h>
> 
> int main()
> {
>  JNIEnv *env;
>  JavaVM *jvm;
>  JavaVMInitArgs vm_args;
>  jint res;
>  jclass cls;
> 
>  vm_args.version = JNI_VERSION_1_2;
> 
>  JNI_GetDefaultJavaVMInitArgs(&vm_args);
> 
>  /* Create the Java VM */
>  res = JNI_CreateJavaVM(&jvm, (void**)&env, &vm_args);
>  if (res < 0)
>     {
>      fprintf(stderr, "Can't create Java VM\n");
>      return(1);
>     }
> 
>  puts("FindClass:");
>  cls = (*env)->FindClass(env, "java/lang/String");
>  if (cls == 0) {
>      fprintf(stderr, "Can't find hello class\n");
>      return(1);
>  }
>  return 0;
> }
> 
> $ gcj -o bug bug.c
> $ ./bug
> FindClass:
> Abort (core dumped)
> 
> $ gcj -v
> Reading specs from
> /data/titan_1/kahlert/gcc-3.1/lib/gcc-lib/sparc-sun-solaris2.7/3.1/specs
> Reading specs from
> /data/titan_1/kahlert/gcc-3.1/lib/gcc-lib/sparc-sun-solaris2.7/3.1/../../../libgcj.spec
> rename spec lib to liborig
> Configured with: ../gcc-20020204/configure --prefix=/data/titan_1/kahlert/gcc-3.1
>    --enable-languages=c++,f77,java --with-as=/usr/ccs/bin/as --with-ld=/usr/ccs/bin/ld
> Thread model: posix
> gcc version 3.1 20020202 (experimental)
> 
> The prog works (i.e. it doesn't segfault) on Linux.
> 
> This is, what i get with dbx ./bug core:
> 
> =>[1] __sigprocmask(0x0, 0xffbeed30, 0x0, 0xffffffff, 0xffffffff, 0x0), at 0xfebd9d18
>   [2] _resetsig(0xfebec9ac, 0x0, 0x0, 0x0, 0x210bc, 0x210c0), at 0xfebcf354
>   [3] _sigon(0xfebf42c0, 0xfebf42a0, 0x210b8, 0xffbeee04, 0x6, 0xfea4df94), at 0xfebceb10
>   [4] _thrp_kill(0x0, 0x1, 0x6, 0xfebec9ac, 0x21040, 0x9), at 0xfebd18e4
>   [5] abort(0xfeab5ad4, 0xfee83160, 0xff154a80, 0x0, 0x0, 0x0), at 0xfea395d0
>   [6] _Jv_ThrowSignal(0xfe984fa8, 0xff1c9f50, 0x0, 0x0, 0x0, 0x0), at 0xfee66984
>   [7] _Z10catch_segviP7siginfoPv(0xb, 0xffbef448, 0xffbef190, 0xfebec9ac, 0x210c8, 0x210a8), at 0xfee669c4
>   [8] 0x0(0x25b68, 0xff1564dc, 0x0, 0xff1700b8, 0x0, 0x0), at 0xffffffffffffffff
>   [9] _Z10wrap_valueIN4java4lang5ClassEEPT_P10_Jv_JNIEnvS4_(0x25b68, 0xff1564dc, 0xfe983140, 0xfee8d49c, 0xffbef608, 0xff0000), at 0xfee6b3f8
>   [10] _Z17_Jv_JNI_FindClassP10_Jv_JNIEnvPKc(0xfe96c118, 0x10ca8, 0xfee6938c, 0x20f1e, 0xfeab5ad4, 0xfeab9708), at 0xfee69478
>   [11] main(0x1, 0xffbef72c, 0xffbef734, 0x20ef8, 0x0, 0x0), at 0x10b90
> 
> Stack level 8 seems strange to me. Shouldn't this be _Jv_JNI_NewLocalRef (env, (jobject) value))?
> 
> Thanks a lot in advance,
> Martin.
> 
> --
> The early bird catches the worm. If you want something else for
> breakfast, get up later.

-- 
Rodrigo Augusto Barbato Ferreira
{M|B}.S. Computer Science, U{NICAMP|FMG}/BRAZIL
Visit the Jewel VM Project at http://www.jewelvm.com/