Hi,
Whenever I tried to set “ba w4 address” for more than 4 breakpoints then it
fails by saying
“Too many data breakpoints for processor 0”
My system is uniproc. What should I do so that I can set more than 4 break
points?
Thanks,
Hakim
Use a different breakpoint type. You are trying to set a breakpoint on
access, meaning a breakpoint when a memory address is read or written.
That’s a hardware register on the CPU, not an inserted int3. There are
only 4 of those registers, so that’s all you get.
What kind of problem are you trying to debug that you need to break on 4
memory addresses at once?
Phil
Philip D. Barila
Seagate Technology LLC
(720) 684-1842
-----Original Message-----
From: xxxxx@lists.osr.com
[mailto:xxxxx@lists.osr.com] On Behalf Of Hakim
Sent: Wednesday, November 16, 2005 1:37 PM
To: Kernel Debugging Interest List
Subject: [windbg] ba
Hi,
Whenever I tried to set “ba w4 address” for more than 4 breakpoints then
it
fails by saying
“Too many data breakpoints for processor 0”
My system is uniproc. What should I do so that I can set more than 4 break
points?
Thanks,
Hakim
You are currently subscribed to windbg as: xxxxx@seagate.com
To unsubscribe send a blank email to xxxxx@lists.osr.com
You are perhaps limited by the hardware registers. Ba’s are set for all
processors AFAIK.
-----Original Message-----
From: xxxxx@lists.osr.com
[mailto:xxxxx@lists.osr.com] On Behalf Of Hakim
Sent: Wednesday, November 16, 2005 12:37 PM
To: Kernel Debugging Interest List
Subject: [windbg] ba
Hi,
Whenever I tried to set “ba w4 address” for more than 4 breakpoints then
it fails by saying
“Too many data breakpoints for processor 0”
My system is uniproc. What should I do so that I can set more than 4
break points?
Thanks,
Hakim
You are currently subscribed to windbg as: xxxxx@appstream.com To
unsubscribe send a blank email to xxxxx@lists.osr.com
Thanks for the informative reply.
Then I have to test 4 at a time.
The driver has 9 device objects and I need to find some state information by
ba w4 address “k;g”
Hakim
wrote in message news:xxxxx@windbg…
> Use a different breakpoint type. You are trying to set a breakpoint on
> access, meaning a breakpoint when a memory address is read or written.
> That’s a hardware register on the CPU, not an inserted int3. There are
> only 4 of those registers, so that’s all you get.
>
> What kind of problem are you trying to debug that you need to break on 4
> memory addresses at once?
>
> Phil
>
> Philip D. Barila
> Seagate Technology LLC
> (720) 684-1842
>
> -----Original Message-----
> From: xxxxx@lists.osr.com
> [mailto:xxxxx@lists.osr.com] On Behalf Of Hakim
> Sent: Wednesday, November 16, 2005 1:37 PM
> To: Kernel Debugging Interest List
> Subject: [windbg] ba
>
> Hi,
>
> Whenever I tried to set “ba w4 address” for more than 4 breakpoints then
> it
> fails by saying
>
> “Too many data breakpoints for processor 0”
>
> My system is uniproc. What should I do so that I can set more than 4 break
>
> points?
>
> Thanks,
> Hakim
>
>
>
> —
> You are currently subscribed to windbg as: xxxxx@seagate.com
> To unsubscribe send a blank email to xxxxx@lists.osr.com
>
>
That sounds like an extremely brute force approach to knowing the “state”
of your object.
What problem are you actually trying to solve that you think you need to
dump the stack every time one of your objects is modified? Are you trying
to find out where something is scribbling garbage on your objects?
Phil
Philip D. Barila
Seagate Technology LLC
(720) 684-1842
-----Original Message-----
From: xxxxx@lists.osr.com
[mailto:xxxxx@lists.osr.com] On Behalf Of Hakim
Sent: Wednesday, November 16, 2005 2:14 PM
To: Kernel Debugging Interest List
Subject: Re:[windbg] ba
Thanks for the informative reply.
Then I have to test 4 at a time.
The driver has 9 device objects and I need to find some state information
by
ba w4 address “k;g”
Hakim
wrote in message news:xxxxx@windbg…
> Use a different breakpoint type. You are trying to set a breakpoint on
> access, meaning a breakpoint when a memory address is read or written.
> That’s a hardware register on the CPU, not an inserted int3. There are
> only 4 of those registers, so that’s all you get.
>
> What kind of problem are you trying to debug that you need to break on 4
> memory addresses at once?
>
> Phil
>
> Philip D. Barila
> Seagate Technology LLC
> (720) 684-1842
>
> -----Original Message-----
> From: xxxxx@lists.osr.com
> [mailto:xxxxx@lists.osr.com] On Behalf Of Hakim
> Sent: Wednesday, November 16, 2005 1:37 PM
> To: Kernel Debugging Interest List
> Subject: [windbg] ba
>
> Hi,
>
> Whenever I tried to set “ba w4 address” for more than 4 breakpoints then
> it
> fails by saying
>
> “Too many data breakpoints for processor 0”
>
> My system is uniproc. What should I do so that I can set more than 4
break
>
> points?
>
> Thanks,
> Hakim
>
>
>
> —
> You are currently subscribed to windbg as: xxxxx@seagate.com
> To unsubscribe send a blank email to xxxxx@lists.osr.com
>
>
—
You are currently subscribed to windbg as: xxxxx@seagate.com
To unsubscribe send a blank email to xxxxx@lists.osr.com
The driver fails to unload during some tests, I need to find which object is
offending by looking at reference count.
Hakim
wrote in message news:xxxxx@windbg…
> That sounds like an extremely brute force approach to knowing the “state”
> of your object.
>
> What problem are you actually trying to solve that you think you need to
> dump the stack every time one of your objects is modified? Are you trying
> to find out where something is scribbling garbage on your objects?
>
> Phil
>
> Philip D. Barila
> Seagate Technology LLC
> (720) 684-1842
>
> -----Original Message-----
> From: xxxxx@lists.osr.com
> [mailto:xxxxx@lists.osr.com] On Behalf Of Hakim
> Sent: Wednesday, November 16, 2005 2:14 PM
> To: Kernel Debugging Interest List
> Subject: Re:[windbg] ba
>
> Thanks for the informative reply.
> Then I have to test 4 at a time.
> The driver has 9 device objects and I need to find some state information
> by
> ba w4 address “k;g”
>
> Hakim
>
>
> wrote in message news:xxxxx@windbg…
>> Use a different breakpoint type. You are trying to set a breakpoint on
>> access, meaning a breakpoint when a memory address is read or written.
>> That’s a hardware register on the CPU, not an inserted int3. There are
>> only 4 of those registers, so that’s all you get.
>>
>> What kind of problem are you trying to debug that you need to break on 4
>> memory addresses at once?
>>
>> Phil
>>
>> Philip D. Barila
>> Seagate Technology LLC
>> (720) 684-1842
>>
>> -----Original Message-----
>> From: xxxxx@lists.osr.com
>> [mailto:xxxxx@lists.osr.com] On Behalf Of Hakim
>> Sent: Wednesday, November 16, 2005 1:37 PM
>> To: Kernel Debugging Interest List
>> Subject: [windbg] ba
>>
>> Hi,
>>
>> Whenever I tried to set “ba w4 address” for more than 4 breakpoints then
>> it
>> fails by saying
>>
>> “Too many data breakpoints for processor 0”
>>
>> My system is uniproc. What should I do so that I can set more than 4
> break
>>
>> points?
>>
>> Thanks,
>> Hakim
>>
>>
>>
>> —
>> You are currently subscribed to windbg as: xxxxx@seagate.com
>> To unsubscribe send a blank email to xxxxx@lists.osr.com
>>
>>
>
>
>
> —
> You are currently subscribed to windbg as: xxxxx@seagate.com
> To unsubscribe send a blank email to xxxxx@lists.osr.com
>
>
Hakim wrote:
The driver fails to unload during some tests, I need to find which object is
offending by looking at reference count.
Wouldn’t it be more efficient to KdPrint a snapshot of the reference
count and the address of the device object every time you either
increase or decrease the reference count?
–
Tim Roberts, xxxxx@probo.com
Providenza & Boekelheide, Inc.
Not really, since OS does most ref\deref.
Hakim
“Tim Roberts” wrote in message news:xxxxx@windbg…
> Hakim wrote:
>
>>The driver fails to unload during some tests, I need to find which object
>>is offending by looking at reference count.
>>
>
> Wouldn’t it be more efficient to KdPrint a snapshot of the reference count
> and the address of the device object every time you either increase or
> decrease the reference count?
>
> –
> Tim Roberts, xxxxx@probo.com
> Providenza & Boekelheide, Inc.
>
>
>
You can still get the reference count, and the bug is likely to be in your
ref\deref’s not the OS’es.
–
Don Burn (MVP, Windows DDK)
Windows 2k/XP/2k3 Filesystem and Driver Consulting
Remove StopSpam from the email to reply
“Hakim” wrote in message news:xxxxx@windbg…
> Not really, since OS does most ref\deref.
>
> Hakim
>
> “Tim Roberts” wrote in message news:xxxxx@windbg…
>> Hakim wrote:
>>
>>>The driver fails to unload during some tests, I need to find which object
>>>is offending by looking at reference count.
>>>
>>
>> Wouldn’t it be more efficient to KdPrint a snapshot of the reference
>> count and the address of the device object every time you either increase
>> or decrease the reference count?
>>
>> –
>> Tim Roberts, xxxxx@probo.com
>> Providenza & Boekelheide, Inc.
>>
>>
>>
>
>
>
Oh Sure, it’s in my driver. I just need to narrow it down for the mistmatch
ref\deref.
Hakim
“Don Burn” wrote in message news:xxxxx@windbg…
> You can still get the reference count, and the bug is likely to be in your
> ref\deref’s not the OS’es.
>
>
> –
> Don Burn (MVP, Windows DDK)
> Windows 2k/XP/2k3 Filesystem and Driver Consulting
> Remove StopSpam from the email to reply
>
>
>
> “Hakim” wrote in message news:xxxxx@windbg…
>> Not really, since OS does most ref\deref.
>>
>> Hakim
>>
>> “Tim Roberts” wrote in message news:xxxxx@windbg…
>>> Hakim wrote:
>>>
>>>>The driver fails to unload during some tests, I need to find which
>>>>object is offending by looking at reference count.
>>>>
>>>
>>> Wouldn’t it be more efficient to KdPrint a snapshot of the reference
>>> count and the address of the device object every time you either
>>> increase or decrease the reference count?
>>>
>>> –
>>> Tim Roberts, xxxxx@probo.com
>>> Providenza & Boekelheide, Inc.
>>>
>>>
>>>
>>
>>
>>
>
>
>