You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@harmony.apache.org by "Vladimir Beliaev (JIRA)" <ji...@apache.org> on 2007/12/13 15:31:44 UTC
[jira] Created: (HARMONY-5305) [drlvm][EUT] regression in jdtdebug
in Linux x86_64
[drlvm][EUT] regression in jdtdebug in Linux x86_64
---------------------------------------------------
Key: HARMONY-5305
URL: https://issues.apache.org/jira/browse/HARMONY-5305
Project: Harmony
Issue Type: Bug
Components: DRLVM
Environment: Linux x86_64
Reporter: Vladimir Beliaev
The jdtdebug suite start hanging on Linux x86_64.
The issue migt be the same to HARMONY-5303, still CC start repoting this crash earlier - the latest non-crashed revision is 595048 (15-Nov-2007)
http://people.apache.org/~smishura/r595048/Linux_x86_64/eut33/results/html/org.eclipse.jdt.debug.tests_linux.gtk.x86_64.html
See the run instruction in HARMONY-5303, still use jdtdebug instead of coreresources.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (HARMONY-5305) [drlvm][jit][EUT] regression in
jdtdebug in Linux x86_64
Posted by "Gregory Shimansky (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HARMONY-5305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12551943 ]
Gregory Shimansky commented on HARMONY-5305:
--------------------------------------------
EUT jdtdebug passes with an exception for testStepIntoSourceMethod test which fails with an interactive message. So the whole suite may be also considered as hanging. Why this test fails I still didn't figure out.
> [drlvm][jit][EUT] regression in jdtdebug in Linux x86_64
> --------------------------------------------------------
>
> Key: HARMONY-5305
> URL: https://issues.apache.org/jira/browse/HARMONY-5305
> Project: Harmony
> Issue Type: Bug
> Components: DRLVM
> Environment: Linux x86_64
> Reporter: Vladimir Beliaev
> Attachments: Breakpoints.java, IBreakpoints.java
>
>
> The jdtdebug suite start hanging on Linux x86_64.
> The issue migt be the same to HARMONY-5303, still CC start repoting this crash earlier - the latest non-crashed revision is 595048 (15-Nov-2007)
> http://people.apache.org/~smishura/r595048/Linux_x86_64/eut33/results/html/org.eclipse.jdt.debug.tests_linux.gtk.x86_64.html
> See the run instruction in HARMONY-5303, still use jdtdebug instead of coreresources.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (HARMONY-5305) [drlvm][jit][EUT] regression in
jdtdebug in Linux x86_64
Posted by "Gregory Shimansky (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HARMONY-5305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12551942 ]
Gregory Shimansky commented on HARMONY-5305:
--------------------------------------------
Yes, EUT no longer hang so this patch helps with passing the tests. Also the AIOOBE exception that I saw dumped on console several times when running EUT jdtdebug no longer appears so it seems like the same bug with index is fixed somewhere else too.
> [drlvm][jit][EUT] regression in jdtdebug in Linux x86_64
> --------------------------------------------------------
>
> Key: HARMONY-5305
> URL: https://issues.apache.org/jira/browse/HARMONY-5305
> Project: Harmony
> Issue Type: Bug
> Components: DRLVM
> Environment: Linux x86_64
> Reporter: Vladimir Beliaev
> Attachments: Breakpoints.java, IBreakpoints.java
>
>
> The jdtdebug suite start hanging on Linux x86_64.
> The issue migt be the same to HARMONY-5303, still CC start repoting this crash earlier - the latest non-crashed revision is 595048 (15-Nov-2007)
> http://people.apache.org/~smishura/r595048/Linux_x86_64/eut33/results/html/org.eclipse.jdt.debug.tests_linux.gtk.x86_64.html
> See the run instruction in HARMONY-5303, still use jdtdebug instead of coreresources.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (HARMONY-5305) [drlvm][jit][EUT] regression in
jdtdebug in Linux x86_64
Posted by "George Timoshenko (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HARMONY-5305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12552620 ]
George Timoshenko commented on HARMONY-5305:
--------------------------------------------
Gregory, you are right. My last note was wrong.
> [drlvm][jit][EUT] regression in jdtdebug in Linux x86_64
> --------------------------------------------------------
>
> Key: HARMONY-5305
> URL: https://issues.apache.org/jira/browse/HARMONY-5305
> Project: Harmony
> Issue Type: Bug
> Components: DRLVM
> Environment: Linux x86_64
> Reporter: Vladimir Beliaev
> Attachments: Breakpoints.java, IBreakpoints.java
>
>
> The jdtdebug suite start hanging on Linux x86_64.
> The issue migt be the same to HARMONY-5303, still CC start repoting this crash earlier - the latest non-crashed revision is 595048 (15-Nov-2007)
> http://people.apache.org/~smishura/r595048/Linux_x86_64/eut33/results/html/org.eclipse.jdt.debug.tests_linux.gtk.x86_64.html
> See the run instruction in HARMONY-5303, still use jdtdebug instead of coreresources.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (HARMONY-5305) [drlvm][jit][EUT] regression in
jdtdebug in Linux x86_64
Posted by "Vladimir Beliaev (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HARMONY-5305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12552247 ]
Vladimir Beliaev commented on HARMONY-5305:
-------------------------------------------
Thanks, Gregory, great evaluation! Could you provide the patch please you wrote about ? Then I can double check the "coreresources" suite with it (HARMONY-5303) - it might help with that resources hang also.
Please also get resolved the hang caused by "testStepIntoSourceMethod".
Thanks
Vladimir Beliaev
> [drlvm][jit][EUT] regression in jdtdebug in Linux x86_64
> --------------------------------------------------------
>
> Key: HARMONY-5305
> URL: https://issues.apache.org/jira/browse/HARMONY-5305
> Project: Harmony
> Issue Type: Bug
> Components: DRLVM
> Environment: Linux x86_64
> Reporter: Vladimir Beliaev
> Attachments: Breakpoints.java, IBreakpoints.java
>
>
> The jdtdebug suite start hanging on Linux x86_64.
> The issue migt be the same to HARMONY-5303, still CC start repoting this crash earlier - the latest non-crashed revision is 595048 (15-Nov-2007)
> http://people.apache.org/~smishura/r595048/Linux_x86_64/eut33/results/html/org.eclipse.jdt.debug.tests_linux.gtk.x86_64.html
> See the run instruction in HARMONY-5303, still use jdtdebug instead of coreresources.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (HARMONY-5305) [drlvm][jit][EUT] regression in
jdtdebug in Linux x86_64
Posted by "Gregory Shimansky (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HARMONY-5305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12552492 ]
Gregory Shimansky commented on HARMONY-5305:
--------------------------------------------
Just to clarify my POV. I don't mean that absence of lazy resolution on x86_64 is a bug. It is just what makes a difference because on x86 lazy resolution is enabled. This makes calls to VM.intern and ClassLoader.loadClass calls transparent to the agent, they are not reported by JVMTI, so JDWP agent doesn't see them at all. For this reason there is no confusion of the agent on x86.
> [drlvm][jit][EUT] regression in jdtdebug in Linux x86_64
> --------------------------------------------------------
>
> Key: HARMONY-5305
> URL: https://issues.apache.org/jira/browse/HARMONY-5305
> Project: Harmony
> Issue Type: Bug
> Components: DRLVM
> Environment: Linux x86_64
> Reporter: Vladimir Beliaev
> Attachments: Breakpoints.java, IBreakpoints.java
>
>
> The jdtdebug suite start hanging on Linux x86_64.
> The issue migt be the same to HARMONY-5303, still CC start repoting this crash earlier - the latest non-crashed revision is 595048 (15-Nov-2007)
> http://people.apache.org/~smishura/r595048/Linux_x86_64/eut33/results/html/org.eclipse.jdt.debug.tests_linux.gtk.x86_64.html
> See the run instruction in HARMONY-5303, still use jdtdebug instead of coreresources.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Resolved: (HARMONY-5305) [drlvm][jit][jet][EUT] regression
in jdtdebug in Linux x86_64
Posted by "Vladimir Beliaev (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HARMONY-5305?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Vladimir Beliaev resolved HARMONY-5305.
---------------------------------------
Resolution: Won't Fix
Ok, there is no this regression anymore since that workaround patch was reverted & related EUT/jtdebug test wwas comittted.
As soon as Lazy Resolution is implemented on x86-64 the "Subtask" JIRA should be set to Resolved and EUT/jtdebug patching should be reverted also.
> [drlvm][jit][jet][EUT] regression in jdtdebug in Linux x86_64
> -------------------------------------------------------------
>
> Key: HARMONY-5305
> URL: https://issues.apache.org/jira/browse/HARMONY-5305
> Project: Harmony
> Issue Type: Bug
> Components: DRLVM
> Environment: Linux / x86_64
> Reporter: Vladimir Beliaev
> Assignee: Gregory Shimansky
> Priority: Minor
> Attachments: Breakpoints.java, IBreakpoints.java
>
>
> The jdtdebug suite start hanging on Linux x86_64.
> The issue migt be the same to HARMONY-5303, still CC start repoting this crash earlier - the latest non-crashed revision is 595048 (15-Nov-2007)
> http://people.apache.org/~smishura/r595048/Linux_x86_64/eut33/results/html/org.eclipse.jdt.debug.tests_linux.gtk.x86_64.html
> See the run instruction in HARMONY-5303, still use jdtdebug instead of coreresources.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (HARMONY-5305) [drlvm][jit][EUT] regression in
jdtdebug in Linux x86_64
Posted by "Vladimir Beliaev (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HARMONY-5305?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Vladimir Beliaev updated HARMONY-5305:
--------------------------------------
Summary: [drlvm][jit][EUT] regression in jdtdebug in Linux x86_64 (was: [drlvm][EUT] regression in jdtdebug in Linux x86_64)
> [drlvm][jit][EUT] regression in jdtdebug in Linux x86_64
> --------------------------------------------------------
>
> Key: HARMONY-5305
> URL: https://issues.apache.org/jira/browse/HARMONY-5305
> Project: Harmony
> Issue Type: Bug
> Components: DRLVM
> Environment: Linux x86_64
> Reporter: Vladimir Beliaev
> Attachments: Breakpoints.java, IBreakpoints.java
>
>
> The jdtdebug suite start hanging on Linux x86_64.
> The issue migt be the same to HARMONY-5303, still CC start repoting this crash earlier - the latest non-crashed revision is 595048 (15-Nov-2007)
> http://people.apache.org/~smishura/r595048/Linux_x86_64/eut33/results/html/org.eclipse.jdt.debug.tests_linux.gtk.x86_64.html
> See the run instruction in HARMONY-5303, still use jdtdebug instead of coreresources.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (HARMONY-5305) [drlvm][jit][jet][EUT] regression in
jdtdebug in Linux x86_64
Posted by "Vladimir Beliaev (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HARMONY-5305?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Vladimir Beliaev updated HARMONY-5305:
--------------------------------------
Summary: [drlvm][jit][jet][EUT] regression in jdtdebug in Linux x86_64 (was: [drlvm][jit][EUT] regression in jdtdebug in Linux x86_64)
> [drlvm][jit][jet][EUT] regression in jdtdebug in Linux x86_64
> -------------------------------------------------------------
>
> Key: HARMONY-5305
> URL: https://issues.apache.org/jira/browse/HARMONY-5305
> Project: Harmony
> Issue Type: Bug
> Components: DRLVM
> Environment: Linux x86_64
> Reporter: Vladimir Beliaev
> Attachments: Breakpoints.java, IBreakpoints.java
>
>
> The jdtdebug suite start hanging on Linux x86_64.
> The issue migt be the same to HARMONY-5303, still CC start repoting this crash earlier - the latest non-crashed revision is 595048 (15-Nov-2007)
> http://people.apache.org/~smishura/r595048/Linux_x86_64/eut33/results/html/org.eclipse.jdt.debug.tests_linux.gtk.x86_64.html
> See the run instruction in HARMONY-5303, still use jdtdebug instead of coreresources.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (HARMONY-5305) [drlvm][jit][EUT] regression in
jdtdebug in Linux x86_64
Posted by "Gregory Shimansky (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HARMONY-5305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12552286 ]
Gregory Shimansky commented on HARMONY-5305:
--------------------------------------------
I pasted the patch in the end of my comment https://issues.apache.org/jira/browse/HARMONY-5305#action_12551933
But I can't say that it is really a fix. It is a workaround for this particular problem when the array index is a long converted to int. In this case index is "fixed" to really fit into int range. But it doesn't make the index actually correct, and it also doesn't fix other situation when incorrect value is not converted to int. So I can't promise that this patch would actually fix any other bug.
Yes I'm going to take care of testStepIntoSourceMethod.
> [drlvm][jit][EUT] regression in jdtdebug in Linux x86_64
> --------------------------------------------------------
>
> Key: HARMONY-5305
> URL: https://issues.apache.org/jira/browse/HARMONY-5305
> Project: Harmony
> Issue Type: Bug
> Components: DRLVM
> Environment: Linux x86_64
> Reporter: Vladimir Beliaev
> Attachments: Breakpoints.java, IBreakpoints.java
>
>
> The jdtdebug suite start hanging on Linux x86_64.
> The issue migt be the same to HARMONY-5303, still CC start repoting this crash earlier - the latest non-crashed revision is 595048 (15-Nov-2007)
> http://people.apache.org/~smishura/r595048/Linux_x86_64/eut33/results/html/org.eclipse.jdt.debug.tests_linux.gtk.x86_64.html
> See the run instruction in HARMONY-5303, still use jdtdebug instead of coreresources.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (HARMONY-5305) [drlvm][jit][EUT] regression in
jdtdebug in Linux x86_64
Posted by "Vladimir Beliaev (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HARMONY-5305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12552343 ]
Vladimir Beliaev commented on HARMONY-5305:
-------------------------------------------
Hi, George, the original issue is opened against SLES10, so I believe Gregory is using this one. So most probable his environment is:
$ cat /etc/SuSE-release
SUSE Linux Enterprise Server 10 (x86_64)
VERSION = 10
$ gcc --version
gcc (GCC) 4.1.0 (SUSE Linux)
> [drlvm][jit][EUT] regression in jdtdebug in Linux x86_64
> --------------------------------------------------------
>
> Key: HARMONY-5305
> URL: https://issues.apache.org/jira/browse/HARMONY-5305
> Project: Harmony
> Issue Type: Bug
> Components: DRLVM
> Environment: Linux x86_64
> Reporter: Vladimir Beliaev
> Attachments: Breakpoints.java, IBreakpoints.java
>
>
> The jdtdebug suite start hanging on Linux x86_64.
> The issue migt be the same to HARMONY-5303, still CC start repoting this crash earlier - the latest non-crashed revision is 595048 (15-Nov-2007)
> http://people.apache.org/~smishura/r595048/Linux_x86_64/eut33/results/html/org.eclipse.jdt.debug.tests_linux.gtk.x86_64.html
> See the run instruction in HARMONY-5303, still use jdtdebug instead of coreresources.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (HARMONY-5305) [drlvm][jit][EUT] regression in
jdtdebug in Linux x86_64
Posted by "Gregory Shimansky (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HARMONY-5305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12551933 ]
Gregory Shimansky commented on HARMONY-5305:
--------------------------------------------
I think I've found a fix for the bug with NPE. The method org/eclipse/jdt/core/dom/DefaultCommentMapper.getExtendedEnd(Lorg/eclipse/jdt/core/dom/ASTNode;)I contains the following bytecodes:
74: lload_3
75: l2i
76: aaload
The index for array is long, it is converted to int and then aaload is executed. For it JET generates the following code:
74: lload_3
75: l2i
0x00002aaacbfc3720: sub $0x8,%rsp
0x00002aaacbfc3724: mov %r12,%rdi
0x00002aaacbfc3727: mov %rsi,0xfffffffffffffda8(%rbp)
0x00002aaacbfc372e: mov $0x2aaaada1892c,%r8
0x00002aaacbfc3738: rex64Z callq *%r8 <- execute rt_h_i64_2_i32, result is in RAX
0x00002aaacbfc373b: add $0x8,%rsp
76: aaload
0x00002aaacbfc373f: mov 0xfffffffffffffda8(%rbp),%r9
0x00002aaacbfc3746: mov $0x2aaaae2b8000,%r10
0x00002aaacbfc3750: cmp %r10,%r9
0x00002aaacbfc3753: jne,pt 0x2aaacbfc3790 <- Check for managed null
0x00002aaacbfc375a: mov %eax,0xfffffffffffffda0(%rbp)
0x00002aaacbfc3760: mov %r9,0xfffffffffffffda8(%rbp)
0x00002aaacbfc3767: movl $0x2,0xfffffffffffffdec(%rbp)
0x00002aaacbfc3771: movl $0x1,0xfffffffffffffde0(%rbp)
0x00002aaacbfc377b: sub $0x8,%rsp
0x00002aaacbfc377f: mov $0x2aaaacdc3200,%r11
0x00002aaacbfc3789: rex64Z callq *%r11
0x00002aaacbfc378c: add $0x8,%rsp
0x00002aaacbfc3790: cmp %eax,0x10(%r9)
0x00002aaacbfc3794: ja,pt 0x2aaacbfc37d1 <- Check for array bounds, note EAX is checked, not RAX
0x00002aaacbfc379b: mov %eax,0xfffffffffffffda0(%rbp)
0x00002aaacbfc37a1: mov %r9,0xfffffffffffffda8(%rbp)
0x00002aaacbfc37a8: movl $0x2,0xfffffffffffffdec(%rbp)
0x00002aaacbfc37b2: movl $0x1,0xfffffffffffffde0(%rbp)
0x00002aaacbfc37bc: sub $0x8,%rsp
0x00002aaacbfc37c0: mov $0x2aaaacdc3350,%r11
0x00002aaacbfc37ca: rex64Z callq *%r11
0x00002aaacbfc37cd: add $0x8,%rsp
0x00002aaacbfc37d1: mov 0x18(%r9,%rax,4),%ecx <- SIGSEGV here R9 = 0x2aaab490f328 RAX = 0x100000001
0x00002aaacbfc37d6: mov $0x2aaaae2b8000,%r11
0x00002aaacbfc37e0: add %r11,%rcx
0x00002aaacbfc37e3: orl $0x20,0xfffffffffffffde8(%rbp)
The R9 is base address for array. RAX is its index converted to int with rt_h_i64_2_i32. For index range JET checks EAX since array indexes are int. But for some reason RAX value is 0x100000001 which means that upper half of RAX is not zero.
Looking at the code in rt_h_i64_2_i32 I see the following:
0x00002aaaada1892c <_ZN7Jitrino3Jet14rt_h_i64_2_i32Ex+0>: push %rbp
0x00002aaaada1892d <_ZN7Jitrino3Jet14rt_h_i64_2_i32Ex+1>: mov %rsp,%rbp
0x00002aaaada18930 <_ZN7Jitrino3Jet14rt_h_i64_2_i32Ex+4>: mov %rdi,0xfffffffffffffff8(%rbp)
0x00002aaaada18934 <_ZN7Jitrino3Jet14rt_h_i64_2_i32Ex+8>: mov 0xfffffffffffffff8(%rbp),%rax
0x00002aaaada18938 <_ZN7Jitrino3Jet14rt_h_i64_2_i32Ex+12>: leaveq
0x00002aaaada18939 <_ZN7Jitrino3Jet14rt_h_i64_2_i32Ex+13>: retq
There is no sign of clearing the upper half of long value in this code. So if upper half of it contained garbage, then it remains there. A workaround for this bug is a patch like this:
Index: vm/jitrino/src/jet/arith_rt.cpp
===================================================================
--- vm/jitrino/src/jet/arith_rt.cpp (revision 604244)
+++ vm/jitrino/src/jet/arith_rt.cpp (working copy)
@@ -268,7 +268,7 @@
double __stdcall rt_h_i32_2_dbl(int i) { return (double)i; };
// i64 ->
-int __stdcall rt_h_i64_2_i32(int64 i) { return (int32)i; };
+int __stdcall rt_h_i64_2_i32(int64 i) { return (int32)(i & 0xffffffff); };
float __stdcall rt_h_i64_2_flt(int64 i) { return (float)i; };
double __stdcall rt_h_i64_2_dbl(int64 i) { return (double)i; };
// flt ->
I am not sure if the bug is in the helper or JET, I am inclined to think the latter since looking at the code, the long index couldn't get the value 0x100000001 in normal conditions.
But since my workaround helps with the problem I am going to check EUT now to see whether jdtdebug actually fails because of this problem.
> [drlvm][jit][EUT] regression in jdtdebug in Linux x86_64
> --------------------------------------------------------
>
> Key: HARMONY-5305
> URL: https://issues.apache.org/jira/browse/HARMONY-5305
> Project: Harmony
> Issue Type: Bug
> Components: DRLVM
> Environment: Linux x86_64
> Reporter: Vladimir Beliaev
> Attachments: Breakpoints.java, IBreakpoints.java
>
>
> The jdtdebug suite start hanging on Linux x86_64.
> The issue migt be the same to HARMONY-5303, still CC start repoting this crash earlier - the latest non-crashed revision is 595048 (15-Nov-2007)
> http://people.apache.org/~smishura/r595048/Linux_x86_64/eut33/results/html/org.eclipse.jdt.debug.tests_linux.gtk.x86_64.html
> See the run instruction in HARMONY-5303, still use jdtdebug instead of coreresources.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (HARMONY-5305) [drlvm][jit][EUT] regression in
jdtdebug in Linux x86_64
Posted by "George Timoshenko (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HARMONY-5305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12552364 ]
George Timoshenko commented on HARMONY-5305:
--------------------------------------------
the only note I'd like to make:
instead of
int __stdcall rt_h_i64_2_i32(int64 i) { return (int32)(i & 0xffffffff); };
That turns into:
push %rbp
mov %rsp,%rbp
mov %rdi,0xfffffffffffffff8(%rbp)
mov 0xfffffffffffffff8(%rbp),%rax
mov %eax,%edx
mov $0xffffffff,%eax
and %edx,%eax
leaveq
retq
I'd suggest
int __stdcall rt_h_i64_2_i32(int64 i) { return 0xffffffff & (int32)i; };
The result is exactly what we need:
push %rbp
mov %rsp,%rbp
mov %rdi,0xfffffffffffffff8(%rbp)
mov 0xfffffffffffffff8(%rbp),%rax
leaveq
retq
> [drlvm][jit][EUT] regression in jdtdebug in Linux x86_64
> --------------------------------------------------------
>
> Key: HARMONY-5305
> URL: https://issues.apache.org/jira/browse/HARMONY-5305
> Project: Harmony
> Issue Type: Bug
> Components: DRLVM
> Environment: Linux x86_64
> Reporter: Vladimir Beliaev
> Attachments: Breakpoints.java, IBreakpoints.java
>
>
> The jdtdebug suite start hanging on Linux x86_64.
> The issue migt be the same to HARMONY-5303, still CC start repoting this crash earlier - the latest non-crashed revision is 595048 (15-Nov-2007)
> http://people.apache.org/~smishura/r595048/Linux_x86_64/eut33/results/html/org.eclipse.jdt.debug.tests_linux.gtk.x86_64.html
> See the run instruction in HARMONY-5303, still use jdtdebug instead of coreresources.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Closed: (HARMONY-5305) [drlvm][jit][jet][EUT] regression in
jdtdebug in Linux x86_64
Posted by "Vladimir Beliaev (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HARMONY-5305?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Vladimir Beliaev closed HARMONY-5305.
-------------------------------------
> [drlvm][jit][jet][EUT] regression in jdtdebug in Linux x86_64
> -------------------------------------------------------------
>
> Key: HARMONY-5305
> URL: https://issues.apache.org/jira/browse/HARMONY-5305
> Project: Harmony
> Issue Type: Bug
> Components: DRLVM
> Environment: Linux / x86_64
> Reporter: Vladimir Beliaev
> Assignee: Gregory Shimansky
> Priority: Minor
> Attachments: Breakpoints.java, IBreakpoints.java
>
>
> The jdtdebug suite start hanging on Linux x86_64.
> The issue migt be the same to HARMONY-5303, still CC start repoting this crash earlier - the latest non-crashed revision is 595048 (15-Nov-2007)
> http://people.apache.org/~smishura/r595048/Linux_x86_64/eut33/results/html/org.eclipse.jdt.debug.tests_linux.gtk.x86_64.html
> See the run instruction in HARMONY-5303, still use jdtdebug instead of coreresources.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (HARMONY-5305) [drlvm][jit][EUT] regression in
jdtdebug in Linux x86_64
Posted by "Gregory Shimansky (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HARMONY-5305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12552405 ]
Gregory Shimansky commented on HARMONY-5305:
--------------------------------------------
BTW looking at the assembly code in your version of helper
push %rbp
mov %rsp,%rbp
mov %rdi,0xfffffffffffffff8(%rbp)
mov 0xfffffffffffffff8(%rbp),%rax
leaveq
retq
I don't quite understand the difference with the current code without my patch:
0x00002aaaada1892c <_ZN7Jitrino3Jet14rt_h_i64_2_i32Ex+0>: push %rbp
0x00002aaaada1892d <_ZN7Jitrino3Jet14rt_h_i64_2_i32Ex+1>: mov %rsp,%rbp
0x00002aaaada18930 <_ZN7Jitrino3Jet14rt_h_i64_2_i32Ex+4>: mov %rdi,0xfffffffffffffff8(%rbp)
0x00002aaaada18934 <_ZN7Jitrino3Jet14rt_h_i64_2_i32Ex+8>: mov 0xfffffffffffffff8(%rbp),%rax
0x00002aaaada18938 <_ZN7Jitrino3Jet14rt_h_i64_2_i32Ex+12>: leaveq
0x00002aaaada18939 <_ZN7Jitrino3Jet14rt_h_i64_2_i32Ex+13>: retq
The code seems to be identical to me and should be working in the same way allowing garbage in the upper half of RAX.
> [drlvm][jit][EUT] regression in jdtdebug in Linux x86_64
> --------------------------------------------------------
>
> Key: HARMONY-5305
> URL: https://issues.apache.org/jira/browse/HARMONY-5305
> Project: Harmony
> Issue Type: Bug
> Components: DRLVM
> Environment: Linux x86_64
> Reporter: Vladimir Beliaev
> Attachments: Breakpoints.java, IBreakpoints.java
>
>
> The jdtdebug suite start hanging on Linux x86_64.
> The issue migt be the same to HARMONY-5303, still CC start repoting this crash earlier - the latest non-crashed revision is 595048 (15-Nov-2007)
> http://people.apache.org/~smishura/r595048/Linux_x86_64/eut33/results/html/org.eclipse.jdt.debug.tests_linux.gtk.x86_64.html
> See the run instruction in HARMONY-5303, still use jdtdebug instead of coreresources.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (HARMONY-5305) [drlvm][jit][EUT] regression in
jdtdebug in Linux x86_64
Posted by "Gregory Shimansky (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HARMONY-5305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12552488 ]
Gregory Shimansky commented on HARMONY-5305:
--------------------------------------------
It looks to me that the bug in "testStepIntoSourceMethod" happens because lazy resolution is disabled in JET on x86_64. When StepIntoSelectionClass.step method is invoked, a lot of Java methods are called in while its compilation. JDWP agent seems to be confused by all of the events and doesn't notice when StepIntoSelectionClass.step is actually invoked. JVMTI correctly sends an event to the agent but for some reason Eclipse doesn't see it.
> [drlvm][jit][EUT] regression in jdtdebug in Linux x86_64
> --------------------------------------------------------
>
> Key: HARMONY-5305
> URL: https://issues.apache.org/jira/browse/HARMONY-5305
> Project: Harmony
> Issue Type: Bug
> Components: DRLVM
> Environment: Linux x86_64
> Reporter: Vladimir Beliaev
> Attachments: Breakpoints.java, IBreakpoints.java
>
>
> The jdtdebug suite start hanging on Linux x86_64.
> The issue migt be the same to HARMONY-5303, still CC start repoting this crash earlier - the latest non-crashed revision is 595048 (15-Nov-2007)
> http://people.apache.org/~smishura/r595048/Linux_x86_64/eut33/results/html/org.eclipse.jdt.debug.tests_linux.gtk.x86_64.html
> See the run instruction in HARMONY-5303, still use jdtdebug instead of coreresources.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (HARMONY-5305) [drlvm][jit][jet][EUT] regression
in jdtdebug in Linux x86_64
Posted by "Gregory Shimansky (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HARMONY-5305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12553635 ]
Gregory Shimansky commented on HARMONY-5305:
--------------------------------------------
I've committed my patch at 605882. But I would like to understand how it happened that RAX has garbage in its upper half, so I don't close the bug yet, just lower its priority because EUT suite should work correctly now.
> [drlvm][jit][jet][EUT] regression in jdtdebug in Linux x86_64
> -------------------------------------------------------------
>
> Key: HARMONY-5305
> URL: https://issues.apache.org/jira/browse/HARMONY-5305
> Project: Harmony
> Issue Type: Bug
> Components: DRLVM
> Environment: Linux / x86_64
> Reporter: Vladimir Beliaev
> Attachments: Breakpoints.java, IBreakpoints.java
>
>
> The jdtdebug suite start hanging on Linux x86_64.
> The issue migt be the same to HARMONY-5303, still CC start repoting this crash earlier - the latest non-crashed revision is 595048 (15-Nov-2007)
> http://people.apache.org/~smishura/r595048/Linux_x86_64/eut33/results/html/org.eclipse.jdt.debug.tests_linux.gtk.x86_64.html
> See the run instruction in HARMONY-5303, still use jdtdebug instead of coreresources.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (HARMONY-5305) [drlvm][jit][jet][EUT] regression in
jdtdebug in Linux x86_64
Posted by "Gregory Shimansky (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HARMONY-5305?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Gregory Shimansky updated HARMONY-5305:
---------------------------------------
Priority: Minor (was: Major)
Assignee: Gregory Shimansky
> [drlvm][jit][jet][EUT] regression in jdtdebug in Linux x86_64
> -------------------------------------------------------------
>
> Key: HARMONY-5305
> URL: https://issues.apache.org/jira/browse/HARMONY-5305
> Project: Harmony
> Issue Type: Bug
> Components: DRLVM
> Environment: Linux / x86_64
> Reporter: Vladimir Beliaev
> Assignee: Gregory Shimansky
> Priority: Minor
> Attachments: Breakpoints.java, IBreakpoints.java
>
>
> The jdtdebug suite start hanging on Linux x86_64.
> The issue migt be the same to HARMONY-5303, still CC start repoting this crash earlier - the latest non-crashed revision is 595048 (15-Nov-2007)
> http://people.apache.org/~smishura/r595048/Linux_x86_64/eut33/results/html/org.eclipse.jdt.debug.tests_linux.gtk.x86_64.html
> See the run instruction in HARMONY-5303, still use jdtdebug instead of coreresources.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (HARMONY-5305) [drlvm][jit][EUT] regression in
jdtdebug in Linux x86_64
Posted by "George Timoshenko (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HARMONY-5305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12552356 ]
George Timoshenko commented on HARMONY-5305:
--------------------------------------------
I have reproduced that RAX. (EM64T Linux SLES10, gcc (GCC) 4.1.2)
IMO the fix Gregory suggested is not a workaround for DRLVM. It is a workaround for gcc on SLES10.
So I am considering it as a normal fix for the problem
> [drlvm][jit][EUT] regression in jdtdebug in Linux x86_64
> --------------------------------------------------------
>
> Key: HARMONY-5305
> URL: https://issues.apache.org/jira/browse/HARMONY-5305
> Project: Harmony
> Issue Type: Bug
> Components: DRLVM
> Environment: Linux x86_64
> Reporter: Vladimir Beliaev
> Attachments: Breakpoints.java, IBreakpoints.java
>
>
> The jdtdebug suite start hanging on Linux x86_64.
> The issue migt be the same to HARMONY-5303, still CC start repoting this crash earlier - the latest non-crashed revision is 595048 (15-Nov-2007)
> http://people.apache.org/~smishura/r595048/Linux_x86_64/eut33/results/html/org.eclipse.jdt.debug.tests_linux.gtk.x86_64.html
> See the run instruction in HARMONY-5303, still use jdtdebug instead of coreresources.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (HARMONY-5305) [drlvm][EUT] regression in jdtdebug
in Linux x86_64
Posted by "Gregory Shimansky (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HARMONY-5305?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Gregory Shimansky updated HARMONY-5305:
---------------------------------------
Attachment: Breakpoints.java
IBreakpoints.java
Attached are two sources from EUT test programs. Please paste them as is including all comments (!!) when you create them in eclipse workspace. It seems that the failing class is the one that deals with comments in sources.
> [drlvm][EUT] regression in jdtdebug in Linux x86_64
> ---------------------------------------------------
>
> Key: HARMONY-5305
> URL: https://issues.apache.org/jira/browse/HARMONY-5305
> Project: Harmony
> Issue Type: Bug
> Components: DRLVM
> Environment: Linux x86_64
> Reporter: Vladimir Beliaev
> Attachments: Breakpoints.java, IBreakpoints.java
>
>
> The jdtdebug suite start hanging on Linux x86_64.
> The issue migt be the same to HARMONY-5303, still CC start repoting this crash earlier - the latest non-crashed revision is 595048 (15-Nov-2007)
> http://people.apache.org/~smishura/r595048/Linux_x86_64/eut33/results/html/org.eclipse.jdt.debug.tests_linux.gtk.x86_64.html
> See the run instruction in HARMONY-5303, still use jdtdebug instead of coreresources.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (HARMONY-5305) [drlvm][EUT] regression in
jdtdebug in Linux x86_64
Posted by "Gregory Shimansky (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HARMONY-5305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12551850 ]
Gregory Shimansky commented on HARMONY-5305:
--------------------------------------------
I think I've found out what makes tests hang. It is not a hang, it is an infinite loop in EUT. Eclipse tries to launch Breakpoints test but it is incorrectly compiled after testSingleListenerMoveCompilationUnit. This test tries to copy main type Breakpoints to BreakpointsCopy file. But it seems that on Linux x86_64 there are some problems with this operation. So there remain several class files that have incorrect source file for public class. Eclipse compiles such classes and generates java.lang.Error: The public type Breakpoints must be defined in its own file
I identified a way how to reproduce at least one problem in renaming main java type in Eclipse. Take attached IBreakpoints.java and Breakpoints.java and create corresponding interface and class in workspace. Click on Breakpoints.java in source tree, select Refactor->Rename. Type in new name BreakpointsCopy, don't change any checkboxes, click Finish. When warned about main type click Finish again. Eclipse will show a message that NPE has occurred while doing this operation. I identified that the method which causes NPE is
org/eclipse/jdt/core/dom/DefaultCommentMapper.getExtendedEnd(Lorg/eclipse/jdt/core/dom/ASTNode;)I
I checked this scenario on JET (NPE happens) and on OPT (NPE doesn't happen). So I believe that the bug is in JET.
BTW for some reason NPE has no stack trace so I didn't find the stack which lead to the failing method.
> [drlvm][EUT] regression in jdtdebug in Linux x86_64
> ---------------------------------------------------
>
> Key: HARMONY-5305
> URL: https://issues.apache.org/jira/browse/HARMONY-5305
> Project: Harmony
> Issue Type: Bug
> Components: DRLVM
> Environment: Linux x86_64
> Reporter: Vladimir Beliaev
>
> The jdtdebug suite start hanging on Linux x86_64.
> The issue migt be the same to HARMONY-5303, still CC start repoting this crash earlier - the latest non-crashed revision is 595048 (15-Nov-2007)
> http://people.apache.org/~smishura/r595048/Linux_x86_64/eut33/results/html/org.eclipse.jdt.debug.tests_linux.gtk.x86_64.html
> See the run instruction in HARMONY-5303, still use jdtdebug instead of coreresources.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (HARMONY-5305) [drlvm][EUT] regression in
jdtdebug in Linux x86_64
Posted by "Vladimir Beliaev (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HARMONY-5305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12551797 ]
Vladimir Beliaev commented on HARMONY-5305:
-------------------------------------------
Another way to start the suite:
# prepare resources
unzip eclipse-Automated-Tests-3.3.zip
cd eclipse-testing
unzip eclipse-junit-tests-I20070625-1500.zip
tar xzf ../eclipse-SDK-3.3-linux-gtk-x86_64.tar.gz
# execute the script below:
export JAVA_HOME=<HARMONY_JDK_HOME>
export PATH=$JAVA_HOME/bin:$PATH
${JAVA_HOME}/bin/java \
-showversion \
-jar eclipse/plugins/org.eclipse.equinox.launcher_1.0.0.v20070606.jar \
-application org.eclipse.test.uitestapplication \
-dev bin \
-data workspace \
formatter=org.apache.tools.ant.taskdefs.optional.junit.XMLJUnitResultFormatter,log.xml \
-testpluginname org.eclipse.jdt.debug.tests \
-classname org.eclipse.jdt.debug.tests.AutomatedSuite \
2>&1 | tee log.txt
> [drlvm][EUT] regression in jdtdebug in Linux x86_64
> ---------------------------------------------------
>
> Key: HARMONY-5305
> URL: https://issues.apache.org/jira/browse/HARMONY-5305
> Project: Harmony
> Issue Type: Bug
> Components: DRLVM
> Environment: Linux x86_64
> Reporter: Vladimir Beliaev
>
> The jdtdebug suite start hanging on Linux x86_64.
> The issue migt be the same to HARMONY-5303, still CC start repoting this crash earlier - the latest non-crashed revision is 595048 (15-Nov-2007)
> http://people.apache.org/~smishura/r595048/Linux_x86_64/eut33/results/html/org.eclipse.jdt.debug.tests_linux.gtk.x86_64.html
> See the run instruction in HARMONY-5303, still use jdtdebug instead of coreresources.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (HARMONY-5305) [drlvm][jit][EUT] regression in
jdtdebug in Linux x86_64
Posted by "Gregory Shimansky (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HARMONY-5305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12552398 ]
Gregory Shimansky commented on HARMONY-5305:
--------------------------------------------
I am just curious about the way how garbage happened to be in RAX upper half. Looking at Eclipse code for getExtendedEnd RAX holds the value of "range" variable. As far as I understand it shouldn't have a value of 0x100000001 in any place of the execution.
public int getExtendedEnd(ASTNode node) {
int end = node.getStartPosition() + node.getLength();
if (this.trailingPtr >= 0) {
long range = -1;
for (int i=0; range<0 && i<=this.trailingPtr; i++) {
if (this.trailingNodes[i] == node) range = this.trailingIndexes[i];
}
if (range >= 0) {
Comment lastComment = this.comments[(int) range];
end = lastComment.getStartPosition() + lastComment.getLength();
}
}
return end-1;
}
> [drlvm][jit][EUT] regression in jdtdebug in Linux x86_64
> --------------------------------------------------------
>
> Key: HARMONY-5305
> URL: https://issues.apache.org/jira/browse/HARMONY-5305
> Project: Harmony
> Issue Type: Bug
> Components: DRLVM
> Environment: Linux x86_64
> Reporter: Vladimir Beliaev
> Attachments: Breakpoints.java, IBreakpoints.java
>
>
> The jdtdebug suite start hanging on Linux x86_64.
> The issue migt be the same to HARMONY-5303, still CC start repoting this crash earlier - the latest non-crashed revision is 595048 (15-Nov-2007)
> http://people.apache.org/~smishura/r595048/Linux_x86_64/eut33/results/html/org.eclipse.jdt.debug.tests_linux.gtk.x86_64.html
> See the run instruction in HARMONY-5303, still use jdtdebug instead of coreresources.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (HARMONY-5305) [drlvm][jit][jet][EUT] regression in
jdtdebug in Linux x86_64
Posted by "Vladimir Beliaev (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HARMONY-5305?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Vladimir Beliaev updated HARMONY-5305:
--------------------------------------
Environment: Linux / x86_64 (was: Linux x86_64)
> [drlvm][jit][jet][EUT] regression in jdtdebug in Linux x86_64
> -------------------------------------------------------------
>
> Key: HARMONY-5305
> URL: https://issues.apache.org/jira/browse/HARMONY-5305
> Project: Harmony
> Issue Type: Bug
> Components: DRLVM
> Environment: Linux / x86_64
> Reporter: Vladimir Beliaev
> Attachments: Breakpoints.java, IBreakpoints.java
>
>
> The jdtdebug suite start hanging on Linux x86_64.
> The issue migt be the same to HARMONY-5303, still CC start repoting this crash earlier - the latest non-crashed revision is 595048 (15-Nov-2007)
> http://people.apache.org/~smishura/r595048/Linux_x86_64/eut33/results/html/org.eclipse.jdt.debug.tests_linux.gtk.x86_64.html
> See the run instruction in HARMONY-5303, still use jdtdebug instead of coreresources.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (HARMONY-5305) [drlvm][jit][EUT] regression in
jdtdebug in Linux x86_64
Posted by "George Timoshenko (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HARMONY-5305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12552337 ]
George Timoshenko commented on HARMONY-5305:
--------------------------------------------
Gregory, please clarify the OS and compiler you built DRLVM with.
I have tried on EM64T Linux SLES9 gcc (GCC) 3.3.3.
And I've got :
mov 0xfffffffffffffff8(%rbp),%eax // <-- EAX, not RAX
at <Jitrino::Jet::rt_h_i64_2_i32(long long)+9>
So the conversion goes as it is supposed to go...
> [drlvm][jit][EUT] regression in jdtdebug in Linux x86_64
> --------------------------------------------------------
>
> Key: HARMONY-5305
> URL: https://issues.apache.org/jira/browse/HARMONY-5305
> Project: Harmony
> Issue Type: Bug
> Components: DRLVM
> Environment: Linux x86_64
> Reporter: Vladimir Beliaev
> Attachments: Breakpoints.java, IBreakpoints.java
>
>
> The jdtdebug suite start hanging on Linux x86_64.
> The issue migt be the same to HARMONY-5303, still CC start repoting this crash earlier - the latest non-crashed revision is 595048 (15-Nov-2007)
> http://people.apache.org/~smishura/r595048/Linux_x86_64/eut33/results/html/org.eclipse.jdt.debug.tests_linux.gtk.x86_64.html
> See the run instruction in HARMONY-5303, still use jdtdebug instead of coreresources.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.