If all agree, please submit your change as a patch to that patch :)
Salikh Zakirov wrote:
Pavel Rebriy wrote:
Fix for Harmony-1582 brocks initialization of JVMTI support. I'm
investigating the problem and going to create fixing JIRA as soon as
possible.
The fix is ready. See http://issues.apache.org/jira/browse/HARMONY-1826
Pavel, I have reviewed the patch in HARMONY-1826, and it looks great!
A couple of minor comments.
The change like this doesn't really move us closer to multi-VM support,
as the jni_env is hardcoded to the emitted code.
- s = mov(s, M_Base_Opnd(esp_reg, 0), Imm_Opnd((int)jni_native_intf)) ; //
o0=jni_native_intf
+ s = mov(s, M_Base_Opnd(esp_reg, 0),
Imm_Opnd((int)p_TLS_vmthread->jni_env)) ; // o0=jni_native_intf
The following change seems to fix unrelated issue ('Fixed Single Step event
disabling'?),
so I think it would have been better if it filed as a separate JIRA/patch.
-
ti->vm_brpt->release_intf(vm_thread->ss_state->predicted_breakpoints);
- _deallocate((unsigned char *)vm_thread->ss_state);
- vm_thread->ss_state = NULL;
+ if( vm_thread->ss_state ) {
+ if( vm_thread->ss_state->predicted_breakpoints ) {
+
ti->vm_brpt->release_intf(vm_thread->ss_state->predicted_breakpoints);
+ }
+ _deallocate((unsigned char *)vm_thread->ss_state);
+ vm_thread->ss_state = NULL;
+ }
Anyway, I vote for a quick inclusion of this patch.
---------------------------------------------------------------------
Terms of use : http://incubator.apache.org/harmony/mailing.html
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
---------------------------------------------------------------------
Terms of use : http://incubator.apache.org/harmony/mailing.html
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
