---- Reported by sven@gnuwhv.de 2005-02-21 18:00:31 MST ----

Description of Problem:
I'm trying to install mono-1.1.4. with  ./configure --prefix=/usr/local
--with-jit=yes && make.
The compilation fails in the directory mcs/classMicrosoft.VisualBasic with
the following exception: 
Unhandled Exception: System.ExecutionEngineException: SIGILL
(see the complete error below)




Steps to reproduce the problem:
1. cd mono-1.1.4 ;-)
2. ./configure --prefix=/usr/local --with-jit=yes
3. make

Actual Results:
Unhandled Exception: System.ExecutionEngineException: SIGILL

Expected Results:
a clean compile


How often does this happen? 
every time

Additional Information:
the complete error:
--snippet--
make[8]: Gehe in Verzeichnis
»/home/sk/software/mono-1.1.4/mcs/class/Microsoft.VisualBasic«
MONO_PATH="../../class/lib/default:$MONO_PATH"
/home/sk/software/mono-1.1.4/runtime/mono-wrapper
../../class/lib/default/mcs.exe   -d:NET_1_1 -d:ONLY_1_1 -debug /noconfig
/r:mscorlib.dll /r:System.dll /r:System.Windows.Forms.dll
@Microsoft.VisualBasic.dll.resources -codepage:65001 /target:library
/out:../../class/lib/default/Microsoft.VisualBasic.dll 
@Microsoft.VisualBasic.dll.sources

Unhandled Exception: System.ExecutionEngineException: SIGILL
in (unmanaged) (wrapper managed-to-native)
System.MonoCustomAttrs:GetCustomAttributesInternal
(System.Reflection.ICustomAttributeProvider,bool)
in <0x00090> (wrapper managed-to-native)
System.MonoCustomAttrs:GetCustomAttributesInternal
(System.Reflection.ICustomAttributeProvider,bool)
in <0x00020> System.MonoCustomAttrs:GetCustomAttributesBase
(System.Reflection.ICustomAttributeProvider)
in <0x000a4> System.MonoCustomAttrs:GetCustomAttributes
(System.Reflection.ICustomAttributeProvider,System.Type,bool)
in <0x00038> System.MonoCustomAttrs:GetCustomAttribute
(System.Reflection.ICustomAttributeProvider,System.Type,bool)
in <0x00030> System.Attribute:GetCustomAttribute
(System.Reflection.MemberInfo,System.Type,bool)
in <0x000b4> Mono.CSharp.AttributeTester:GetMemberObsoleteAttribute
(System.Reflection.MemberInfo)
in <0x00130> Mono.CSharp.AttributeTester:GetMethodObsoleteAttribute
(System.Reflection.MethodBase)
in <0x00214> Mono.CSharp.Invocation:EmitCall
(Mono.CSharp.EmitContext,bool,bool,Mono.CSharp.Expression,System.Reflection.MethodBase,System.Collections.ArrayList,Mono.CSharp.Location,bool,bool)
in <0x00050> Mono.CSharp.Invocation:EmitCall
(Mono.CSharp.EmitContext,bool,bool,Mono.CSharp.Expression,System.Reflection.MethodBase,System.Collections.ArrayList,Mono.CSharp.Location)
in <0x00200> Mono.CSharp.PropertyExpr:EmitAssign
(Mono.CSharp.EmitContext,Mono.CSharp.Expression,bool,bool)
in <0x00258> Mono.CSharp.Assign:Emit (Mono.CSharp.EmitContext,bool)
in <0x0002c> Mono.CSharp.Assign:EmitStatement (Mono.CSharp.EmitContext)
in <0x00034> Mono.CSharp.StatementExpression:DoEmit (Mono.CSharp.EmitContext)
in <0x00050> Mono.CSharp.Statement:Emit (Mono.CSharp.EmitContext)
in <0x0012c> Mono.CSharp.Block:DoEmit (Mono.CSharp.EmitContext)
in <0x003b0> Mono.CSharp.Block:Emit (Mono.CSharp.EmitContext)
in <0x0012c> Mono.CSharp.Block:DoEmit (Mono.CSharp.EmitContext)
in <0x003b0> Mono.CSharp.Block:Emit (Mono.CSharp.EmitContext)
in <0x00044> Mono.CSharp.EmitContext:EmitResolvedTopBlock
(Mono.CSharp.ToplevelBlock,bool)
in <0x000ac> Mono.CSharp.EmitContext:EmitTopBlock
(Mono.CSharp.ToplevelBlock,Mono.CSharp.InternalParameters,Mono.CSharp.Location)
in <0x004f0> Mono.CSharp.MethodData:Emit
(Mono.CSharp.TypeContainer,Mono.CSharp.Attributable)
in <0x00050> Mono.CSharp.Method:Emit ()
in <0x006e4> Mono.CSharp.TypeContainer:EmitType ()
in <0x005c0> Mono.CSharp.RootContext:EmitCode ()
in <0x00ea4> Mono.CSharp.Driver:MainDriver (string[])
in <0x00024> Mono.CSharp.Driver:Main (string[])

make[8]: *** [../../class/lib/default/Microsoft.VisualBasic.dll] Fehler 1
make[8]: Verlasse Verzeichnis
»/home/sk/software/mono-1.1.4/mcs/class/Microsoft.VisualBasic«
make[7]: *** [all.real] Fehler 2
make[7]: Verlasse Verzeichnis
»/home/sk/software/mono-1.1.4/mcs/class/Microsoft.VisualBasic«
make[6]: *** [all-recursive] Fehler 1
make[6]: Verlasse Verzeichnis »/home/sk/software/mono-1.1.4/mcs/class«
make[5]: *** [all-recursive] Fehler 1
make[5]: Verlasse Verzeichnis »/home/sk/software/mono-1.1.4/mcs«
make[4]: *** [profile-do--default--all] Fehler 2
make[4]: Verlasse Verzeichnis »/home/sk/software/mono-1.1.4/mcs«
make[3]: *** [profiles-do--all] Fehler 2
make[3]: Verlasse Verzeichnis »/home/sk/software/mono-1.1.4/mcs«
make[2]: *** [all-local] Fehler 2
make[2]: Verlasse Verzeichnis »/home/sk/software/mono-1.1.4/runtime«
make[1]: *** [all-recursive] Fehler 1
make[1]: Verlasse Verzeichnis »/home/sk/software/mono-1.1.4«
make: *** [all] Fehler 2

--snippet--



---- Additional Comments From grompf@sublimeintervention.com 2005-02-21 20:11:50 MST ----

I am unable to replicate this with 1.1.4 or HEAD on my ubuntu/ppc box (hoary)

Whats your processor/machine?

-kangaroo




---- Additional Comments From sven@gnuwhv.de 2005-02-21 21:39:58 MST ----

I'm using an apple-ibook with a G3 processor ...
mono-1.0.6 compiles cleanly on it.  



---- Additional Comments From grompf@sublimeintervention.com 2005-02-22 18:11:22 MST ----

paco has confirmed this bug on his (g3) computer as well.

I cannot confirm this bug on a G4

G4 Output:
MONO_PATH="../../class/lib/default:$MONO_PATH" mono  ../../class/lib/default/mcs.exe   
-d:NET_1_1 -d:ONLY_1_1 -debug /noconfig /r:mscorlib.dll /r:System.dll /r:
System.Windows.Forms.dll @Microsoft.VisualBasic.dll.resources -codepage:65001 /target:
library /out:../../class/lib/default/Microsoft.VisualBasic.dll  
@Microsoft.VisualBasic.dll.sourcesMONO_PATH="../../class/lib/
net_1_1_bootstrap/:$MONO_PATH" mono  ../../class/lib/net_1_1_bootstrap/sn.exe -q -R 
../../class/lib/default/Microsoft.VisualBasic.dll ../../class/mono.snk
Assembly ../../class/lib/default/Microsoft.VisualBasic.dll signed.

Paco indicated that he'll give me ssh access to dig into this bug to his g3.

Additionally a user "EY" from mono appears to have found another testcase that exhibits 
this behaviour on G3 and will attach it here.

-kangaroo

(Updated topic & product to be appropriate)



---- Additional Comments From jeyoung@priscimon.com 2005-02-22 18:12:25 MST ----

Created an attachment (id=167447)
The test CS file.




---- Additional Comments From grompf@sublimeintervention.com 2005-02-23 13:08:03 MST ----

Better testcase:

using System;

class T {
        static void Main (string[] args) {
                throw new Exception ();
        }
}

It looks like exception throwing is broken on the g3;

gdb info:

Program received signal EXC_BAD_INSTRUCTION, Illegal instruction/operand.
0x00141f84 in start.2 ()
(gdb) bt
#0  0x00141f84 in start.2 ()
#1  0x00082a34 in arch_handle_exception (ctx=0xc24bc195, obj=0xc24bc195, 
test_only=-2021289069) at exceptions-ppc.c:893
#2  0x00081958 in throw_exception (exc=0x87859393, eip=2273678227, 
esp=2273678227, int_regs=0xc24bc195, fp_regs=0xc24bc195, rethrow=-1035222635) 
at exceptions-ppc.c:299
#3  0x001421d0 in start.7 ()
#4  0x0077a50c in ?? ()
#5  0x0077a408 in ?? ()

(gdb) p print_method_from_ip (0x0077a50c)
IP 0x77a50c at offset 0x3c of method T:Main (string[]) (0x77a4d0 0x77a520)[domain 
0x48bf20 - test.exe]
$1 = void

(gdb) disas
Dump of assembler code for function start.2:
0x00141f84 <start.2+0>: .long 0x0
0x00141f88 <start.2+4>: .long 0x0
0x00141f8c <start.2+8>: .long 0x0
0x00141f90 <start.2+12>:        .long 0x0
0x00141f94 <start.2+16>:        .long 0x0
<snip>

-kangaroo




---- Additional Comments From sven@gnuwhv.de 2005-02-24 07:54:04 MST ----

mono-1.0.6 compiles cleanly - but when running f-spot on mono-1.0.6
the same error occurs when scanning for Images:

---
Scanning /home/sk/Pictures/Images/Backgrounds
value = 2005:02:24 12:46:08 len = 19

Unhandled Exception: System.ExecutionEngineException: SIGILL
in (unmanaged) (wrapper managed-to-native) Gtk.Application:gtk_main ()
in <0x00080> (wrapper managed-to-native) Gtk.Application:gtk_main ()
in <0x00014> Gtk.Application:Run ()
in <0x00018> Gnome.Program:Run ()
in [0x0007c] (at /home/sk/Software/f-spot-0.0.9/src/main.cs:31)
Driver:Main (string[])
---





---- Additional Comments From jeffrey.knight@gmail.com 2005-03-07 11:20:53 MST ----

I can replicate Kangaroo's test case, and am trying to fix this.



---- Additional Comments From lupus@ximian.com 2005-03-07 11:23:57 MST ----

Note that disas should not be used, especially on osx, to disassemble
code. Use, instead:
x/Ni address
where N is the number of instructions to disassemble.
It's always useful to disassemble a few instructions before
the instruction address as well.



---- Additional Comments From jeffrey.knight@gmail.com 2005-03-18 00:01:14 MST ----

(gdb) t a a bt

Thread 3 (Thread 16386 (LWP 23103)):
#0  0x0fc51a8c in kill () from /lib/libc.so.6
#1  0x0fe3f9e4 in __pthread_restart_new () from /lib/libpthread.so.0
#2  0x0fe417f4 in __pthread_alt_unlock () from /lib/libpthread.so.0
#3  0x0fe417f4 in __pthread_alt_unlock () from /lib/libpthread.so.0
#4  0x0fe417f4 in __pthread_alt_unlock () from /lib/libpthread.so.0
#5  0x0fe417f4 in __pthread_alt_unlock () from /lib/libpthread.so.0
#6  0x0fe417f4 in __pthread_alt_unlock () from /lib/libpthread.so.0
#7  0x0fe417f4 in __pthread_alt_unlock () from /lib/libpthread.so.0
#8  0x0fe417f4 in __pthread_alt_unlock () from /lib/libpthread.so.0
#9  0x0fe417f4 in __pthread_alt_unlock () from /lib/libpthread.so.0
#10 0x0fe417f4 in __pthread_alt_unlock () from /lib/libpthread.so.0
#11 0x0fe417f4 in __pthread_alt_unlock () from /lib/libpthread.so.0
#12 0x0fe417f4 in __pthread_alt_unlock () from /lib/libpthread.so.0
Previous frame inner to this frame (corrupt stack?)
#0  0x1017ccbc in start.3 ()

(gdb) info registers
r0             0x1017ccbc       269995196
r1             0x7ffff080       2147479680
r2             0x3003aa60       805546592
r3             0x7ffff248       2147480136
r4             0x30214414       807486484
r5             0x101d5f30       270360368
r6             0x3c     60
r7             0x48     72
r8             0x24     36
r9             0x0      0
r10            0xfd4c5d0        265602512
r11            0x0      0
r12            0xfc8e840        264824896
r13            0x10182744       270018372
r14            0x40000000       1073741824
r15            0x1      1
r16            0x7ffff0b8       2147479736
r17            0x101e8330       270435120
r18            0x7ffff098       2147479704
r19            0x1      1
r20            0x10180000       270008320
r21            0x101d5f30       270360368
r22            0x2      2
r23            0x101caf00       270315264
r24            0x1018a298       270049944
r25            0x101d5f30       270360368
r26            0x10187108       270037256
r27            0x0      0
r28            0x7ffff248       2147480136
r29            0x0      0
r30            0x10188e98       270044824
r31            0x10188e98       270044824
pc             0x1017ccbc       269995196
ps             0x8d032  577586
cr             0x22444842       574900290
lr             0x100f75c8       269448648
ctr            0x1017ccbc       269995196
xer            0x20000000       536870912
fpscr          0x1036bb84       272022404
vscr           0x0      0
vrsave         0x0      0




---- Additional Comments From lupus@ximian.com 2005-03-22 14:15:04 MST ----

Fixed in svn.
Thanks to Andy Satori for providing access to a G3 box.

Imported an attachment (id=167447)

Unknown bug field "cf_op_sys_details" encountered while moving bug
   <cf_op_sys_details>ubuntu 4.10 ppc</cf_op_sys_details>