Re: TMTask and 10.4, 10.4.1 - please help!
- Subject: Re: TMTask and 10.4, 10.4.1 - please help!
- From: Alice Hartley <email@hidden>
- Date: Wed, 1 Jun 2005 01:02:56 -0400
- Delivered-to: email@hidden
- Delivered-to: email@hidden
Can anyone tell me what has changed in 10.4 in TMTask.
The relevant code has worked without problem since
Feb 2003 until the advent of 10.4.
The crash happens seldom and I don't really know where to begin
to look for the cause or solution.
Thanks
-----------------------
Here is some of the relevent code.
CARBON is defined, PMCL_OSX_NATIVE_KERNEL is defined.
void
cmain_vbl (TMTaskPtr task)
{
#ifdef PMCL_OSX_NATIVE_KERNEL
mach_port_t
current_thread_on_entry = TScurrent_mach_thread,
suspended_thread = (mach_port_t)0;
#endif
lock_binding_stack();
#ifdef PMCL_OSX_NATIVE_KERNEL
if ((current_thread_on_entry != ((mach_port_t) 0)) &&
(current_thread_on_entry != mach_thread_self())) {
if (thread_suspend(current_thread_on_entry) == KERN_SUCCESS) {
suspended_thread = current_thread_on_entry;
}
}
#endif
{
struct lispsymbol *interrupt_level_sym = &nrs_INTERRUPT_LEVEL;
int interrupt_level = (int) interrupt_level_sym->vcell;
LispObj tagged_il_sym = ((LispObj) interrupt_level_sym) + fulltag_misc;
if (interrupt_level == 0) {
nrs_INTERRUPT_LEVEL.vcell = 1<<fixnumshift;
}
else if ((tag_of(interrupt_level) == tag_fixnum) && /* This test may not be necessary */
(interrupt_level < 0)) {
struct special_binding *b;
LispObj b_value;
for (b = (special_binding *) lisp_global(DB_LINK); b != 0; b = b->link) {
if (((b->sym) == (lispsymbol *) tagged_il_sym)) {
if ((b_value = (b->value)) == 0) { /* this is line 679 */
b->value = 1<<fixnumshift;
break;
}
else if (b_value == 1<<fixnumshift) {
break;
}
}
}
}
}
#ifdef PMCL_OSX_NATIVE_KERNEL
if (suspended_thread != ((mach_port_t) 0)) {
thread_resume(suspended_thread);
}
#endif
unlock_binding_stack();
/* Reenable the VBL */
#ifndef CARBON
task->vblCount = 1;
#else
PrimeTime((QElemPtr)task, 15);
#endif
}
void
lock_binding_stack()
{
#ifdef PMCL_OSX_NATIVE_KERNEL
lock_acquire(binding_stack_lock_set,0);
#endif
}
void
unlock_binding_stack()
{
#ifdef PMCL_OSX_NATIVE_KERNEL
lock_release(binding_stack_lock_set,0);
#endif
}
#ifdef PMCL_OSX_NATIVE_KERNEL
lock_set_t binding_stack_lock_set;
mach_port_t TScurrent_mach_thread = (mach_port_t) 0;
#endif
>> 0 pmcl-OSX-kernel 0x0069e410 cmain_vbl + 276
>> (pmcl-kernel.c:679)
>
>The crash seems to have occurred inside your cmain_vbl code, while
>accessing memory through r0 or r2. What's going on in that code,
>near line 679 of pmcl-kernel.c?
>
>--Dave
>
>On May 26, 2005, at 1:48 PM, Alice Hartley wrote:
>
>> Users of our application report the following crash:
>>
>> Exception: EXC_BAD_ACCESS (0x0001)
>> Codes: KERN_INVALID_ADDRESS (0x0001) at 0x81061505
>>
>>
>> Thread 2 Crashed:
>> 0 pmcl-OSX-kernel 0x0069e410 cmain_vbl + 276
>> (pmcl-kernel.c:679)
>> 1 ...ple.CoreServices.CarbonCore 0x90b738b4 TimerThread + 288
>> 2 libSystem.B.dylib 0x9002c3d4 _pthread_body + 96
>>
>>
>>
>> Thread 2 crashed with PPC Thread State:
>> srr0: 0x0069e410 srr1: 0x0000d030 vrsave: 0x00000000
>> cr: 0x20000008 xer: 0x00000000 lr: 0x0069e338 ctr: 0x9000a770
>> r0: 0x81061501 r1: 0xf0101d70 r2: 0x81061501 r3: 0x00000000
>> r4: 0x00000003 r5: 0x00000024 r6: 0x0000002c r7: 0x00002407
>> r8: 0x00000000 r9: 0x00000000 r10: 0x89d832be r11: 0xa0006c68
>> r12: 0x9000a770 r13: 0x00000000 r14: 0x00000000 r15: 0x00000000
>> r16: 0x00000000 r17: 0x00000000 r18: 0x00000000 r19: 0x00000000
>> r20: 0x00000000 r21: 0x00000000 r22: 0x00000000 r23: 0x00000000
>> r24: 0x00000000 r25: 0x00000000 r26: 0x00000000 r27: 0x00000000
>> r28: 0xa0b037a4 r29: 0x0069e2fc r30: 0xf0101d70 r31: 0x0069e314
>>
>> Another report:
>>
>>>> Exception: EXC_BAD_ACCESS (0x0001)
>>>> Codes: KERN_INVALID_ADDRESS (0x0001) at 0x81041505
>>>>
>>>> Thread 2 Crashed:
>>>> 0 pmcl-OSX-kernel 0x00621410 cmain_vbl + 276
>>>> (pmcl-kernel.c:660)
>>>> 1 ...ple.CoreServices.CarbonCore 0x90b738ac TimerThread + 288
>>>> 2 libSystem.B.dylib 0x9002c3b4 _pthread_body
>>>> + 96
>>>>
>>>> Thread 2 crashed with PPC Thread State:
>>>> srr0: 0x00621410 srr1: 0x0000d030 vrsave:
>>>> 0x00000000
>>>> cr: 0x80000002 xer: 0x00000000 lr: 0x00621338 ctr:
>>>> 0x9000a750
>>>> r0: 0x81041501 r1: 0xf0101d70 r2: 0x00000000 r3:
>>>> 0x00000000
>>>> r4: 0x00000003 r5: 0x00000024 r6: 0x0000002c r7:
>>>> 0x00002407
>>>> r8: 0x00000000 r9: 0x81041501 r10: 0xb0f66ec0 r11:
>>>> 0xa0006c68
>>>> r12: 0x9000a750 r13: 0x00000000 r14: 0x00000000 r15:
>>>> 0x00000000
>>>> r16: 0x00000000 r17: 0x00000000 r18: 0x00000000 r19:
>>>> 0x00000000
>>>> r20: 0x00000000 r21: 0x00000000 r22: 0x00000000 r23:
>>>> 0x00000000
>>>> r24: 0x00000000 r25: 0x00000000 r26: 0x00000000 r27:
>>>> 0x00000000
>>>> r28: 0xa0b0379c r29: 0x006212fc r30: 0xf0101d70 r31:
>>>> 0x00621314
>>>>
>>>>
>>
>> The crash occurs seldom, on the order of once a day. More likely to
>> happen after the machine has been idle for some time (several hours).
>>
>> Rumor has it that something changed in the implementation of TMTask
>> in 10.4
>>
>> Here is the code that sets it up.
>>
>> {
>> TMTask *q = (TMTask *) NewPtrClear(sizeof(TMTask));
>> /* cmain_descriptor = NewTimerProc(cmain_vbl); */
>> cmain_descriptor = (TimerUPP)NewTimerUPP(cmain_vbl);
>>
>> q->qType = 0; /* ?? */
>> q->tmAddr = cmain_descriptor;
>> q->tmCount = 1;
>> cmain_vblQ = q;
>> InsTime((QElem *) q);
>> PrimeTime((QElem *) q, 15);
>> }
>>
>> Any help would be much appreciated.
>> Thank you.
>>
>>
>> _______________________________________________
>> Do not post admin requests to the list. They will be ignored.
>> Carbon-dev mailing list (email@hidden)
>> Help/Unsubscribe/Update your Subscription:
>> http://lists.apple.com/mailman/options/carbon-dev/email@hidden
>>
>> This email sent to email@hidden
>>
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Carbon-dev mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/carbon-dev/email@hidden
This email sent to email@hidden
Visit the Apple Store online or at retail locations.
1-800-MY-APPLE
Contact Apple | Terms of Use | Privacy Policy
Copyright © 2007 Apple Inc. All rights reserved.
