Hello,
We have installed Vista (Release Build 6000) on 3 PC .On two PC DebugView each second
get message " Will not retry; Sense/ASC/ASCQ of 02/3a/00" .
I found this text in the cdrom.sys driver . Why driver send this message. How can I disable it?Stab.
Thanks in advice
Mark
Check classpnp and cdrom sources in DDK/src/storage.
“Mark Shnaider” wrote in message news:xxxxx@ntdev…
Hello,
We have installed Vista (Release Build 6000) on 3 PC .On two PC DebugView
each second
get message " Will not retry; Sense/ASC/ASCQ of 02/3a/00" .
I found this text in the cdrom.sys driver . Why driver send this message.
How can I disable it?Stab.
Thanks in advice
Mark
Good idea. To my surprise WDK (both 5600 and 6000) contain sources with it. src\storage\class\cdrom\sense.c calls DbgPrintEx() directly without #if DBG. Shame on the developer who did it. Not because forgot it in the code which is quite understandable but because called the function directly and not via a macro. Which is the only way how to make sure release build doesn’t contains forgotten debug prints.
It isn’t the first and the only case at Vista. I’m affraid it is result of adding DbgPrintEx() which doesn’t make too much sense, IMO. WinDbg won’t display info level prints by default and it is easy to forget them in the code. DebugView displays them all which is probably because it hooks debug print API and not the routine which really prints.
Next unwanted result of DbgPrintEx() is that incompetent support persons don’t see debug prints in WinDbg and close reports about them as “not reproducible”. Even when log and information about using DebugView was included. And DebugView is now MS tool 
An idea: what if DTM test warn about DbgPrint/Ex imports? It checks for unwanted imports anyway (at least HCT did).
Best regards,
Michal Vodicka
UPEK, Inc.
[xxxxx@upek.com, http://www.upek.com]
From: xxxxx@lists.osr.com[SMTP:xxxxx@lists.osr.com] on behalf of cristalink[SMTP:xxxxx@cristalink.com]
Reply To: Windows System Software Devs Interest List
Sent: Wednesday, December 06, 2006 8:14 PM
To: Windows System Software Devs Interest List
Subject: Re:[ntdev] Strange DebugOutStrung from cdrom driverCheck classpnp and cdrom sources in DDK/src/storage.
“Mark Shnaider” wrote in message news:xxxxx@ntdev…
> Hello,
> We have installed Vista (Release Build 6000) on 3 PC .On two PC DebugView
> each second
> get message " Will not retry; Sense/ASC/ASCQ of 02/3a/00" .
> I found this text in the cdrom.sys driver . Why driver send this message.
> How can I disable it?Stab.
> Thanks in advice
> Mark
>
>
>
>
>
>
>
>
>
>
> —
> Questions? First check the Kernel Driver FAQ at http://www.osronline.com/article.cfm?id=256
>
> To unsubscribe, visit the List Server section of OSR Online at http://www.osronline.com/page.cfm?name=ListServer
>
>>To my surprise WDK (both 5600 and 6000) contain sources with it.
They have been there since at least NT4.
“Michal Vodicka” wrote in message
news:xxxxx@ntdev…
Good idea. To my surprise WDK (both 5600 and 6000) contain sources with it.
src\storage\class\cdrom\sense.c calls DbgPrintEx() directly without #if DBG.
Shame on the developer who did it. Not because forgot it in the code which
is quite understandable but because called the function directly and not via
a macro. Which is the only way how to make sure release build doesn’t
contains forgotten debug prints.
It isn’t the first and the only case at Vista. I’m affraid it is result of
adding DbgPrintEx() which doesn’t make too much sense, IMO. WinDbg won’t
display info level prints by default and it is easy to forget them in the
code. DebugView displays them all which is probably because it hooks debug
print API and not the routine which really prints.
Next unwanted result of DbgPrintEx() is that incompetent support persons
don’t see debug prints in WinDbg and close reports about them as “not
reproducible”. Even when log and information about using DebugView was
included. And DebugView is now MS tool
An idea: what if DTM test warn about DbgPrint/Ex imports? It checks for
unwanted imports anyway (at least HCT did).
Best regards,
Michal Vodicka
UPEK, Inc.
[xxxxx@upek.com, http://www.upek.com]
> ----------
> From:
> xxxxx@lists.osr.com[SMTP:xxxxx@lists.osr.com]
> on behalf of cristalink[SMTP:xxxxx@cristalink.com]
> Reply To: Windows System Software Devs Interest List
> Sent: Wednesday, December 06, 2006 8:14 PM
> To: Windows System Software Devs Interest List
> Subject: Re:[ntdev] Strange DebugOutStrung from cdrom driver
>
> Check classpnp and cdrom sources in DDK/src/storage.
>
>
>
> “Mark Shnaider” wrote in message news:xxxxx@ntdev…
> Hello,
> We have installed Vista (Release Build 6000) on 3 PC .On two PC
> DebugView
> each second
> get message " Will not retry; Sense/ASC/ASCQ of 02/3a/00" .
> I found this text in the cdrom.sys driver . Why driver send this message.
> How can I disable it?Stab.
> Thanks in advice
> Mark
>
>
>
>
>
>
>
>
>
>
> —
> Questions? First check the Kernel Driver FAQ at
> http://www.osronline.com/article.cfm?id=256
>
> To unsubscribe, visit the List Server section of OSR Online at
> http://www.osronline.com/page.cfm?name=ListServer
>
I knew CDROM drivers were in DDK/WDK I was surpirsed by a fact there are up-to-date sources which contain this nonsense.
OP can now report problem and point to exact lines of code which should be fixed. However, I don’t expect it’d persuade support persons problem even exists…
Best regards,
Michal Vodicka
UPEK, Inc.
[xxxxx@upek.com, http://www.upek.com]
From: xxxxx@lists.osr.com[SMTP:xxxxx@lists.osr.com] on behalf of cristalink[SMTP:xxxxx@cristalink.com]
Reply To: Windows System Software Devs Interest List
Sent: Wednesday, December 06, 2006 9:07 PM
To: Windows System Software Devs Interest List
Subject: Re:[ntdev] Strange DebugOutStrung from cdrom driver>>To my surprise WDK (both 5600 and 6000) contain sources with it.
They have been there since at least NT4.
“Michal Vodicka” wrote in message
> news:xxxxx@ntdev…
> Good idea. To my surprise WDK (both 5600 and 6000) contain sources with it.
> src\storage\class\cdrom\sense.c calls DbgPrintEx() directly without #if DBG.
> Shame on the developer who did it. Not because forgot it in the code which
> is quite understandable but because called the function directly and not via
> a macro. Which is the only way how to make sure release build doesn’t
> contains forgotten debug prints.
>
> It isn’t the first and the only case at Vista. I’m affraid it is result of
> adding DbgPrintEx() which doesn’t make too much sense, IMO. WinDbg won’t
> display info level prints by default and it is easy to forget them in the
> code. DebugView displays them all which is probably because it hooks debug
> print API and not the routine which really prints.
>
> Next unwanted result of DbgPrintEx() is that incompetent support persons
> don’t see debug prints in WinDbg and close reports about them as “not
> reproducible”. Even when log and information about using DebugView was
> included. And DebugView is now MS tool
>
> An idea: what if DTM test warn about DbgPrint/Ex imports? It checks for
> unwanted imports anyway (at least HCT did).
>
> Best regards,
>
> Michal Vodicka
> UPEK, Inc.
> [xxxxx@upek.com, http://www.upek.com]
>
>
> > ----------
> > From:
> > xxxxx@lists.osr.com[SMTP:xxxxx@lists.osr.com]
> > on behalf of cristalink[SMTP:xxxxx@cristalink.com]
> > Reply To: Windows System Software Devs Interest List
> > Sent: Wednesday, December 06, 2006 8:14 PM
> > To: Windows System Software Devs Interest List
> > Subject: Re:[ntdev] Strange DebugOutStrung from cdrom driver
> >
> > Check classpnp and cdrom sources in DDK/src/storage.
> >
> >
> >
> > “Mark Shnaider” wrote in message news:xxxxx@ntdev…
> > Hello,
> > We have installed Vista (Release Build 6000) on 3 PC .On two PC
> > DebugView
> > each second
> > get message " Will not retry; Sense/ASC/ASCQ of 02/3a/00" .
> > I found this text in the cdrom.sys driver . Why driver send this message.
> > How can I disable it?Stab.
> > Thanks in advice
> > Mark
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> > —
> > Questions? First check the Kernel Driver FAQ at
> > http://www.osronline.com/article.cfm?id=256
> >
> > To unsubscribe, visit the List Server section of OSR Online at
> > http://www.osronline.com/page.cfm?name=ListServer
> >
>
>
>
> —
> Questions? First check the Kernel Driver FAQ at http://www.osronline.com/article.cfm?id=256
>
> To unsubscribe, visit the List Server section of OSR Online at http://www.osronline.com/page.cfm?name=ListServer
>
I doubt it’s something we’ll issue a QFE for. However I’ve told the
storage team about it … they’ll add it to the list of things to
cleanup in whatever the next version of CDROM is.
In the mean time they’ve suggested putting a CD in the drive, as they
suspect the error is coming back because the drive is empty. Putting
something in there should stop the errors and keep it from filling up
your viewer.
-p
-----Original Message-----
From: xxxxx@lists.osr.com
[mailto:xxxxx@lists.osr.com] On Behalf Of Michal Vodicka
Sent: Wednesday, December 06, 2006 12:16 PM
To: Windows System Software Devs Interest List
Subject: RE: [ntdev] Strange DebugOutStrung from cdrom driver
I knew CDROM drivers were in DDK/WDK I was surpirsed by a fact there
are up-to-date sources which contain this nonsense.
OP can now report problem and point to exact lines of code which should
be fixed. However, I don’t expect it’d persuade support persons problem
even exists…
Best regards,
Michal Vodicka
UPEK, Inc.
[xxxxx@upek.com, http://www.upek.com]
From:
xxxxx@lists.osr.com[SMTP:xxxxx@lists.osr.com
] on behalf of cristalink[SMTP:xxxxx@cristalink.com]
Reply To: Windows System Software Devs Interest List
Sent: Wednesday, December 06, 2006 9:07 PM
To: Windows System Software Devs Interest List
Subject: Re:[ntdev] Strange DebugOutStrung from cdrom driver>>To my surprise WDK (both 5600 and 6000) contain sources with it.
They have been there since at least NT4.
“Michal Vodicka” wrote in message
> news:xxxxx@ntdev…
> Good idea. To my surprise WDK (both 5600 and 6000) contain sources
with it.
> src\storage\class\cdrom\sense.c calls DbgPrintEx() directly without
#if DBG.
> Shame on the developer who did it. Not because forgot it in the code
which
> is quite understandable but because called the function directly and
not via
> a macro. Which is the only way how to make sure release build doesn’t
> contains forgotten debug prints.
>
> It isn’t the first and the only case at Vista. I’m affraid it is
result of
> adding DbgPrintEx() which doesn’t make too much sense, IMO. WinDbg
won’t
> display info level prints by default and it is easy to forget them in
the
> code. DebugView displays them all which is probably because it hooks
debug
> print API and not the routine which really prints.
>
> Next unwanted result of DbgPrintEx() is that incompetent support
persons
> don’t see debug prints in WinDbg and close reports about them as “not
> reproducible”. Even when log and information about using DebugView was
> included. And DebugView is now MS tool
>
> An idea: what if DTM test warn about DbgPrint/Ex imports? It checks
for
> unwanted imports anyway (at least HCT did).
>
> Best regards,
>
> Michal Vodicka
> UPEK, Inc.
> [xxxxx@upek.com, http://www.upek.com]
>
>
> > ----------
> > From:
> >
xxxxx@lists.osr.com[SMTP:xxxxx@lists.osr.com
]
> > on behalf of cristalink[SMTP:xxxxx@cristalink.com]
> > Reply To: Windows System Software Devs Interest List
> > Sent: Wednesday, December 06, 2006 8:14 PM
> > To: Windows System Software Devs Interest List
> > Subject: Re:[ntdev] Strange DebugOutStrung from cdrom driver
> >
> > Check classpnp and cdrom sources in DDK/src/storage.
> >
> >
> >
> > “Mark Shnaider” wrote in message news:xxxxx@ntdev…
> > Hello,
> > We have installed Vista (Release Build 6000) on 3 PC .On two PC
> > DebugView
> > each second
> > get message " Will not retry; Sense/ASC/ASCQ of 02/3a/00" .
> > I found this text in the cdrom.sys driver . Why driver send this
message.
> > How can I disable it?Stab.
> > Thanks in advice
> > Mark
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> > —
> > Questions? First check the Kernel Driver FAQ at
> > http://www.osronline.com/article.cfm?id=256
> >
> > To unsubscribe, visit the List Server section of OSR Online at
> > http://www.osronline.com/page.cfm?name=ListServer
> >
>
>
>
> —
> Questions? First check the Kernel Driver FAQ at
http://www.osronline.com/article.cfm?id=256
>
> To unsubscribe, visit the List Server section of OSR Online at
http://www.osronline.com/page.cfm?name=ListServer
>
—
Questions? First check the Kernel Driver FAQ at
http://www.osronline.com/article.cfm?id=256
To unsubscribe, visit the List Server section of OSR Online at
http://www.osronline.com/page.cfm?name=ListServer
Thanks, at least this way works Sorry, I’m a bit angry about it because reported several such problems during beta phase and was dissatisfied with reports handling.
The next possibility is to fix it locally, build driver and replace installed version. Is the WDK 6000 version equivallent with driver supplied with Vista RTM?
Best regards,
Michal Vodicka
UPEK, Inc.
[xxxxx@upek.com, http://www.upek.com]
From: xxxxx@lists.osr.com[SMTP:xxxxx@lists.osr.com] on behalf of Peter Wieland[SMTP:xxxxx@windows.microsoft.com]
Reply To: Windows System Software Devs Interest List
Sent: Wednesday, December 06, 2006 10:36 PM
To: Windows System Software Devs Interest List
Subject: RE: [ntdev] Strange DebugOutStrung from cdrom driverI doubt it’s something we’ll issue a QFE for. However I’ve told the
storage team about it … they’ll add it to the list of things to
cleanup in whatever the next version of CDROM is.In the mean time they’ve suggested putting a CD in the drive, as they
suspect the error is coming back because the drive is empty. Putting
something in there should stop the errors and keep it from filling up
your viewer.-p
-----Original Message-----
From: xxxxx@lists.osr.com
[mailto:xxxxx@lists.osr.com] On Behalf Of Michal Vodicka
Sent: Wednesday, December 06, 2006 12:16 PM
To: Windows System Software Devs Interest List
Subject: RE: [ntdev] Strange DebugOutStrung from cdrom driverI knew CDROM drivers were in DDK/WDK
are up-to-date sources which contain this nonsense.OP can now report problem and point to exact lines of code which should
be fixed. However, I don’t expect it’d persuade support persons problem
even exists…Best regards,
Michal Vodicka
UPEK, Inc.
[xxxxx@upek.com, http://www.upek.com]> ----------
> From:
xxxxx@lists.osr.com[SMTP:xxxxx@lists.osr.com
] on behalf of cristalink[SMTP:xxxxx@cristalink.com]
> Reply To: Windows System Software Devs Interest List
> Sent: Wednesday, December 06, 2006 9:07 PM
> To: Windows System Software Devs Interest List
> Subject: Re:[ntdev] Strange DebugOutStrung from cdrom driver
>
> >>To my surprise WDK (both 5600 and 6000) contain sources with it.
>
> They have been there since at least NT4.
>
>
>
>
> “Michal Vodicka” wrote in message
> > news:xxxxx@ntdev…
> > Good idea. To my surprise WDK (both 5600 and 6000) contain sources
> with it.
> > src\storage\class\cdrom\sense.c calls DbgPrintEx() directly without
> #if DBG.
> > Shame on the developer who did it. Not because forgot it in the code
> which
> > is quite understandable but because called the function directly and
> not via
> > a macro. Which is the only way how to make sure release build doesn’t
> > contains forgotten debug prints.
> >
> > It isn’t the first and the only case at Vista. I’m affraid it is
> result of
> > adding DbgPrintEx() which doesn’t make too much sense, IMO. WinDbg
> won’t
> > display info level prints by default and it is easy to forget them in
> the
> > code. DebugView displays them all which is probably because it hooks
> debug
> > print API and not the routine which really prints.
> >
> > Next unwanted result of DbgPrintEx() is that incompetent support
> persons
> > don’t see debug prints in WinDbg and close reports about them as “not
> > reproducible”. Even when log and information about using DebugView was
>
> > included. And DebugView is now MS tool
> >
> > An idea: what if DTM test warn about DbgPrint/Ex imports? It checks
> for
> > unwanted imports anyway (at least HCT did).
> >
> > Best regards,
> >
> > Michal Vodicka
> > UPEK, Inc.
> > [xxxxx@upek.com, http://www.upek.com]
> >
> > >
> > > ----------
> > > From:
> > >
> xxxxx@lists.osr.com[SMTP:xxxxx@lists.osr.com
>]
> > > on behalf of cristalink[SMTP:xxxxx@cristalink.com]
> > > Reply To: Windows System Software Devs Interest List
> > > Sent: Wednesday, December 06, 2006 8:14 PM
> > > To: Windows System Software Devs Interest List
> > > Subject: Re:[ntdev] Strange DebugOutStrung from cdrom driver
> > >
> > > Check classpnp and cdrom sources in DDK/src/storage.
> > >
> > >
> > >
> > > “Mark Shnaider” wrote in message news:xxxxx@ntdev…
> > > Hello,
> > > We have installed Vista (Release Build 6000) on 3 PC .On two PC
> > > DebugView
> > > each second
> > > get message " Will not retry; Sense/ASC/ASCQ of 02/3a/00" .
> > > I found this text in the cdrom.sys driver . Why driver send this
> message.
> > > How can I disable it?Stab.
> > > Thanks in advice
> > > Mark
> > >
> > >
> > >
> > >
> > >
> > >
> > >
> > >
> > >
> > >
> > > —
> > > Questions? First check the Kernel Driver FAQ at
> > > http://www.osronline.com/article.cfm?id=256
> > >
> > > To unsubscribe, visit the List Server section of OSR Online at
> > > http://www.osronline.com/page.cfm?name=ListServer
> > >
> >
> >
> >
> > —
> > Questions? First check the Kernel Driver FAQ at
> http://www.osronline.com/article.cfm?id=256
> >
> > To unsubscribe, visit the List Server section of OSR Online at
> http://www.osronline.com/page.cfm?name=ListServer
> >
>
> —
> Questions? First check the Kernel Driver FAQ at
> http://www.osronline.com/article.cfm?id=256
>
> To unsubscribe, visit the List Server section of OSR Online at
> http://www.osronline.com/page.cfm?name=ListServer
>
>
> —
> Questions? First check the Kernel Driver FAQ at http://www.osronline.com/article.cfm?id=256
>
> To unsubscribe, visit the List Server section of OSR Online at http://www.osronline.com/page.cfm?name=ListServer
>
In the grand scope of things it allows us to spend our resources (dev &
test) fixing big issues.
However if it’s any consolation it also annoys us when we can’t fix this
piddly stuff too.
-p
-----Original Message-----
From: xxxxx@lists.osr.com
[mailto:xxxxx@lists.osr.com] On Behalf Of Michal Vodicka
Sent: Wednesday, December 06, 2006 2:10 PM
To: Windows System Software Devs Interest List
Subject: RE: [ntdev] Strange DebugOutStrung from cdrom driver
Thanks, at least this way works Sorry, I’m a bit angry about it
because reported several such problems during beta phase and was
dissatisfied with reports handling.
The next possibility is to fix it locally, build driver and replace
installed version. Is the WDK 6000 version equivallent with driver
supplied with Vista RTM?
Best regards,
Michal Vodicka
UPEK, Inc.
[xxxxx@upek.com, http://www.upek.com]
From:
xxxxx@lists.osr.com[SMTP:xxxxx@lists.osr.com
] on behalf of Peter Wieland[SMTP:xxxxx@windows.microsoft.com]
Reply To: Windows System Software Devs Interest List
Sent: Wednesday, December 06, 2006 10:36 PM
To: Windows System Software Devs Interest List
Subject: RE: [ntdev] Strange DebugOutStrung from cdrom driverI doubt it’s something we’ll issue a QFE for. However I’ve told the
storage team about it … they’ll add it to the list of things to
cleanup in whatever the next version of CDROM is.In the mean time they’ve suggested putting a CD in the drive, as they
suspect the error is coming back because the drive is empty. Putting
something in there should stop the errors and keep it from filling up
your viewer.-p
-----Original Message-----
From: xxxxx@lists.osr.com
[mailto:xxxxx@lists.osr.com] On Behalf Of Michal Vodicka
Sent: Wednesday, December 06, 2006 12:16 PM
To: Windows System Software Devs Interest List
Subject: RE: [ntdev] Strange DebugOutStrung from cdrom driverI knew CDROM drivers were in DDK/WDK
there
are up-to-date sources which contain this nonsense.OP can now report problem and point to exact lines of code which
should
be fixed. However, I don’t expect it’d persuade support persons
problem
even exists…Best regards,
Michal Vodicka
UPEK, Inc.
[xxxxx@upek.com, http://www.upek.com]> ----------
> From:
xxxxx@lists.osr.com[SMTP:xxxxx@lists.osr.com
] on behalf of cristalink[SMTP:xxxxx@cristalink.com]
> Reply To: Windows System Software Devs Interest List
> Sent: Wednesday, December 06, 2006 9:07 PM
> To: Windows System Software Devs Interest List
> Subject: Re:[ntdev] Strange DebugOutStrung from cdrom driver
>
> >>To my surprise WDK (both 5600 and 6000) contain sources with it.
>
> They have been there since at least NT4.
>
>
>
>
> “Michal Vodicka” wrote in message
> > news:xxxxx@ntdev…
> > Good idea. To my surprise WDK (both 5600 and 6000) contain sources
> with it.
> > src\storage\class\cdrom\sense.c calls DbgPrintEx() directly without
> #if DBG.
> > Shame on the developer who did it. Not because forgot it in the code
> which
> > is quite understandable but because called the function directly and
> not via
> > a macro. Which is the only way how to make sure release build
doesn’t
> > contains forgotten debug prints.
> >
> > It isn’t the first and the only case at Vista. I’m affraid it is
> result of
> > adding DbgPrintEx() which doesn’t make too much sense, IMO. WinDbg
> won’t
> > display info level prints by default and it is easy to forget them
in
> the
> > code. DebugView displays them all which is probably because it hooks
> debug
> > print API and not the routine which really prints.
> >
> > Next unwanted result of DbgPrintEx() is that incompetent support
> persons
> > don’t see debug prints in WinDbg and close reports about them as
“not
> > reproducible”. Even when log and information about using DebugView
was
>
> > included. And DebugView is now MS tool
> >
> > An idea: what if DTM test warn about DbgPrint/Ex imports? It checks
> for
> > unwanted imports anyway (at least HCT did).
> >
> > Best regards,
> >
> > Michal Vodicka
> > UPEK, Inc.
> > [xxxxx@upek.com, http://www.upek.com]
> >
> > >
> > > ----------
> > > From:
> > >
>
xxxxx@lists.osr.com[SMTP:xxxxx@lists.osr.com
>]
> > > on behalf of cristalink[SMTP:xxxxx@cristalink.com]
> > > Reply To: Windows System Software Devs Interest List
> > > Sent: Wednesday, December 06, 2006 8:14 PM
> > > To: Windows System Software Devs Interest List
> > > Subject: Re:[ntdev] Strange DebugOutStrung from cdrom driver
> > >
> > > Check classpnp and cdrom sources in DDK/src/storage.
> > >
> > >
> > >
> > > “Mark Shnaider” wrote in message
news:xxxxx@ntdev…
> > > Hello,
> > > We have installed Vista (Release Build 6000) on 3 PC .On two PC
> > > DebugView
> > > each second
> > > get message " Will not retry; Sense/ASC/ASCQ of 02/3a/00" .
> > > I found this text in the cdrom.sys driver . Why driver send this
> message.
> > > How can I disable it?Stab.
> > > Thanks in advice
> > > Mark
> > >
> > >
> > >
> > >
> > >
> > >
> > >
> > >
> > >
> > >
> > > —
> > > Questions? First check the Kernel Driver FAQ at
> > > http://www.osronline.com/article.cfm?id=256
> > >
> > > To unsubscribe, visit the List Server section of OSR Online at
> > > http://www.osronline.com/page.cfm?name=ListServer
> > >
> >
> >
> >
> > —
> > Questions? First check the Kernel Driver FAQ at
> http://www.osronline.com/article.cfm?id=256
> >
> > To unsubscribe, visit the List Server section of OSR Online at
> http://www.osronline.com/page.cfm?name=ListServer
> >
>
> —
> Questions? First check the Kernel Driver FAQ at
> http://www.osronline.com/article.cfm?id=256
>
> To unsubscribe, visit the List Server section of OSR Online at
> http://www.osronline.com/page.cfm?name=ListServer
>
>
> —
> Questions? First check the Kernel Driver FAQ at
http://www.osronline.com/article.cfm?id=256
>
> To unsubscribe, visit the List Server section of OSR Online at
http://www.osronline.com/page.cfm?name=ListServer
>
—
Questions? First check the Kernel Driver FAQ at
http://www.osronline.com/article.cfm?id=256
To unsubscribe, visit the List Server section of OSR Online at
http://www.osronline.com/page.cfm?name=ListServer
I quite understand problem with resources and priorities, we have it, too. In my experience the worst problem is when developer doesn’t even know there is a problem and learn support persons if they can’t reproduce a problem, it doesn’t mean it doesn’t exist. In this case it is one minute fix for a tired developer who doesn’t dare to do anything more complicated this day. In the beta phase and if it is logged into bug tracking database even as the lowest priority problem. After release everything is much harder.
BTW, in such cases open source projects have a big advantage. A volunteer makes a fix and developer or project manager only approves it.
Best regards,
Michal Vodicka
UPEK, Inc.
[xxxxx@upek.com, http://www.upek.com]
From: xxxxx@lists.osr.com[SMTP:xxxxx@lists.osr.com] on behalf of Peter Wieland[SMTP:xxxxx@windows.microsoft.com]
Reply To: Windows System Software Devs Interest List
Sent: Wednesday, December 06, 2006 11:21 PM
To: Windows System Software Devs Interest List
Subject: RE: [ntdev] Strange DebugOutStrung from cdrom driverIn the grand scope of things it allows us to spend our resources (dev &
test) fixing big issues.However if it’s any consolation it also annoys us when we can’t fix this
piddly stuff too.-p
-----Original Message-----
From: xxxxx@lists.osr.com
[mailto:xxxxx@lists.osr.com] On Behalf Of Michal Vodicka
Sent: Wednesday, December 06, 2006 2:10 PM
To: Windows System Software Devs Interest List
Subject: RE: [ntdev] Strange DebugOutStrung from cdrom driverThanks, at least this way works
because reported several such problems during beta phase and was
dissatisfied with reports handling.The next possibility is to fix it locally, build driver and replace
installed version. Is the WDK 6000 version equivallent with driver
supplied with Vista RTM?Best regards,
Michal Vodicka
UPEK, Inc.
[xxxxx@upek.com, http://www.upek.com]> ----------
> From:
xxxxx@lists.osr.com[SMTP:xxxxx@lists.osr.com
] on behalf of Peter Wieland[SMTP:xxxxx@windows.microsoft.com]
> Reply To: Windows System Software Devs Interest List
> Sent: Wednesday, December 06, 2006 10:36 PM
> To: Windows System Software Devs Interest List
> Subject: RE: [ntdev] Strange DebugOutStrung from cdrom driver
>
> I doubt it’s something we’ll issue a QFE for. However I’ve told the
> storage team about it … they’ll add it to the list of things to
> cleanup in whatever the next version of CDROM is.
>
> In the mean time they’ve suggested putting a CD in the drive, as they
> suspect the error is coming back because the drive is empty. Putting
> something in there should stop the errors and keep it from filling up
> your viewer.
>
> -p
>
> -----Original Message-----
> From: xxxxx@lists.osr.com
> [mailto:xxxxx@lists.osr.com] On Behalf Of Michal Vodicka
> Sent: Wednesday, December 06, 2006 12:16 PM
> To: Windows System Software Devs Interest List
> Subject: RE: [ntdev] Strange DebugOutStrung from cdrom driver
>
> I knew CDROM drivers were in DDK/WDK
there
> are up-to-date sources which contain this nonsense.
>
> OP can now report problem and point to exact lines of code which
should
> be fixed. However, I don’t expect it’d persuade support persons
problem
> even exists…
>
> Best regards,
>
> Michal Vodicka
> UPEK, Inc.
> [xxxxx@upek.com, http://www.upek.com]
>
>
> > ----------
> > From:
>
xxxxx@lists.osr.com[SMTP:xxxxx@lists.osr.com
> ] on behalf of cristalink[SMTP:xxxxx@cristalink.com]
> > Reply To: Windows System Software Devs Interest List>
> > Sent: Wednesday, December 06, 2006 9:07 PM
> > To: Windows System Software Devs Interest List
> > Subject: Re:[ntdev] Strange DebugOutStrung from cdrom driver
> >
> > >>To my surprise WDK (both 5600 and 6000) contain sources with it.
> >
> > They have been there since at least NT4.
> >
> >
> >
> >
> > “Michal Vodicka” wrote in message
> > > news:xxxxx@ntdev…
> > > Good idea. To my surprise WDK (both 5600 and 6000) contain sources
> > with it.
> > > src\storage\class\cdrom\sense.c calls DbgPrintEx() directly without
> > #if DBG.
> > > Shame on the developer who did it. Not because forgot it in the code
> > which
> > > is quite understandable but because called the function directly and
> > not via
> > > a macro. Which is the only way how to make sure release build
> doesn’t
> > > contains forgotten debug prints.
> > >
> > > It isn’t the first and the only case at Vista. I’m affraid it is
> > result of
> > > adding DbgPrintEx() which doesn’t make too much sense, IMO. WinDbg
> > won’t
> > > display info level prints by default and it is easy to forget them
> in
> > the
> > > code. DebugView displays them all which is probably because it hooks
> > debug
> > > print API and not the routine which really prints.
> > >
> > > Next unwanted result of DbgPrintEx() is that incompetent support
> > persons
> > > don’t see debug prints in WinDbg and close reports about them as
> “not
> > > reproducible”. Even when log and information about using DebugView
> was
> >
> > > included. And DebugView is now MS tool
> > >
> > > An idea: what if DTM test warn about DbgPrint/Ex imports? It checks
> > for
> > > unwanted imports anyway (at least HCT did).
> > >
> > > Best regards,
> > >
> > > Michal Vodicka
> > > UPEK, Inc.
> > > [xxxxx@upek.com, http://www.upek.com]
> > >
> > > >
> > > > ----------
> > > > From:
> > > >
> >
> xxxxx@lists.osr.com[SMTP:xxxxx@lists.osr.com
> >]
> > > > on behalf of cristalink[SMTP:xxxxx@cristalink.com]
> > > > Reply To: Windows System Software Devs Interest List
> > > > Sent: Wednesday, December 06, 2006 8:14 PM
> > > > To: Windows System Software Devs Interest List
> > > > Subject: Re:[ntdev] Strange DebugOutStrung from cdrom driver
> > > >
> > > > Check classpnp and cdrom sources in DDK/src/storage.
> > > >
> > > >
> > > >
> > > > “Mark Shnaider” wrote in message
> news:xxxxx@ntdev…
> > > > Hello,
> > > > We have installed Vista (Release Build 6000) on 3 PC .On two PC
> > > > DebugView
> > > > each second
> > > > get message " Will not retry; Sense/ASC/ASCQ of 02/3a/00" .
> > > > I found this text in the cdrom.sys driver . Why driver send this
> > message.
> > > > How can I disable it?Stab.
> > > > Thanks in advice
> > > > Mark
> > > >
> > > >
> > > >
> > > >
> > > >
> > > >
> > > >
> > > >
> > > >
> > > >
> > > > —
> > > > Questions? First check the Kernel Driver FAQ at
> > > > http://www.osronline.com/article.cfm?id=256
> > > >
> > > > To unsubscribe, visit the List Server section of OSR Online at
> > > > http://www.osronline.com/page.cfm?name=ListServer
> > > >
> > >
> > >
> > >
> > > —
> > > Questions? First check the Kernel Driver FAQ at
> > http://www.osronline.com/article.cfm?id=256
> > >
> > > To unsubscribe, visit the List Server section of OSR Online at
> > http://www.osronline.com/page.cfm?name=ListServer
> > >
> >
> > —
> > Questions? First check the Kernel Driver FAQ at
> > http://www.osronline.com/article.cfm?id=256
> >
> > To unsubscribe, visit the List Server section of OSR Online at
> > http://www.osronline.com/page.cfm?name=ListServer
> >
> >
> > —
> > Questions? First check the Kernel Driver FAQ at
> http://www.osronline.com/article.cfm?id=256
> >
> > To unsubscribe, visit the List Server section of OSR Online at
> http://www.osronline.com/page.cfm?name=ListServer
> >
>
> —
> Questions? First check the Kernel Driver FAQ at>
> http://www.osronline.com/article.cfm?id=256
>
> To unsubscribe, visit the List Server section of OSR Online at
> http://www.osronline.com/page.cfm?name=ListServer
>
>
> —
> Questions? First check the Kernel Driver FAQ at http://www.osronline.com/article.cfm?id=256
>
> To unsubscribe, visit the List Server section of OSR Online at http://www.osronline.com/page.cfm?name=ListServer
>
First off, the developer who made this tragic error was me. I did it.
I’m to blame. Sorry.
Let’s scope the impact of this problem:
- Debugging FRE builds, this is not seen by default. (using kernel
debugger of course) - Even Debugging in CHK builds, it isn’t seen by default.
- It adds a few strings to the binary image; nothing sensitive.
- The only way to have it show up is to enable “DPFLTR_CDROM_ID” to
warning or higher level.
You are correct that this should not have been checked in w/o the use of
macros. As a consolation, Michal, you can now send me up to one flame
mail a month for the next year. 
Since you know that we aren’t likely able to modify this, perhaps you
can request that DebugView be updated to respect the component ID and
levels instead of printing everything? That would also make the utility
more useful on CHK builds…
.
-----Original Message-----
From: Michal Vodicka [mailto:xxxxx@upek.com]
Sent: Wednesday, December 06, 2006 2:49 PM
Subject: RE: Strange DebugOutStrung from cdrom driver
I quite understand problem with resources and priorities, we have it,
too. In my experience the worst problem is when developer doesn’t even
know there is a problem and learn support persons if they can’t
reproduce a problem, it doesn’t mean it doesn’t exist. In this case it
is one minute fix for a tired developer who doesn’t dare to do anything
more complicated this day. In the beta phase and if it is logged into
bug tracking database even as the lowest priority problem. After release
everything is much harder.
BTW, in such cases open source projects have a big advantage. A
volunteer makes a fix and developer or project manager only approves it.
Best regards,
Michal Vodicka
UPEK, Inc.
[xxxxx@upek.com, http://www.upek.com]
From:
xxxxx@lists.osr.com[SMTP:xxxxx@lists.osr.com
] on behalf of Peter Wieland[SMTP:xxxxx@windows.microsoft.com]
Reply To: Windows System Software Devs Interest List
Sent: Wednesday, December 06, 2006 11:21 PM
To: Windows System Software Devs Interest List
Subject: RE: [ntdev] Strange DebugOutStrung from cdrom driverIn the grand scope of things it allows us to spend our resources (dev
&
test) fixing big issues.However if it’s any consolation it also annoys us when we can’t fix
this
piddly stuff too.-p
-----Original Message-----
From: xxxxx@lists.osr.com
[mailto:xxxxx@lists.osr.com] On Behalf Of Michal Vodicka
Sent: Wednesday, December 06, 2006 2:10 PM
To: Windows System Software Devs Interest List
Subject: RE: [ntdev] Strange DebugOutStrung from cdrom driverThanks, at least this way works
because reported several such problems during beta phase and was
dissatisfied with reports handling.The next possibility is to fix it locally, build driver and replace
installed version. Is the WDK 6000 version equivallent with driver
supplied with Vista RTM?Best regards,
Michal Vodicka
UPEK, Inc.
[xxxxx@upek.com, http://www.upek.com]> ----------
> From:
xxxxx@lists.osr.com[SMTP:xxxxx@lists.osr.com
] on behalf of Peter Wieland[SMTP:xxxxx@windows.microsoft.com]
> Reply To: Windows System Software Devs Interest List
> Sent: Wednesday, December 06, 2006 10:36 PM
> To: Windows System Software Devs Interest List
> Subject: RE: [ntdev] Strange DebugOutStrung from cdrom driver
>
> I doubt it’s something we’ll issue a QFE for. However I’ve told the
> storage team about it … they’ll add it to the list of things to
> cleanup in whatever the next version of CDROM is.
>
> In the mean time they’ve suggested putting a CD in the drive, as
they
> suspect the error is coming back because the drive is empty.
Putting
> something in there should stop the errors and keep it from filling
up
> your viewer.
>
> -p
>
> -----Original Message-----
> From: xxxxx@lists.osr.com
> [mailto:xxxxx@lists.osr.com] On Behalf Of Michal
Vodicka
> Sent: Wednesday, December 06, 2006 12:16 PM
> To: Windows System Software Devs Interest List
> Subject: RE: [ntdev] Strange DebugOutStrung from cdrom driver
>
> I knew CDROM drivers were in DDK/WDK
there
> are up-to-date sources which contain this nonsense.
>
> OP can now report problem and point to exact lines of code which
should
> be fixed. However, I don’t expect it’d persuade support persons
problem
> even exists…
>
> Best regards,
>
> Michal Vodicka
> UPEK, Inc.
> [xxxxx@upek.com, http://www.upek.com]
>
>
> > ----------
> > From:
>
xxxxx@lists.osr.com[SMTP:xxxxx@lists.osr.com
> ] on behalf of cristalink[SMTP:xxxxx@cristalink.com]
> > Reply To: Windows System Software Devs Interest List>
> > Sent: Wednesday, December 06, 2006 9:07 PM
> > To: Windows System Software Devs Interest List
> > Subject: Re:[ntdev] Strange DebugOutStrung from cdrom driver
> >
> > >>To my surprise WDK (both 5600 and 6000) contain sources with it.
> >
> > They have been there since at least NT4.
> >
> >
> >
> >
> > “Michal Vodicka” wrote in message
> > > news:xxxxx@ntdev…
> > > Good idea. To my surprise WDK (both 5600 and 6000) contain sources
> > with it.
> > > src\storage\class\cdrom\sense.c calls DbgPrintEx() directly
without
> > #if DBG.
> > > Shame on the developer who did it. Not because forgot it in the
code
> > which
> > > is quite understandable but because called the function directly
and
> > not via
> > > a macro. Which is the only way how to make sure release build
> doesn’t
> > > contains forgotten debug prints.
> > >
> > > It isn’t the first and the only case at Vista. I’m affraid it is
> > result of
> > > adding DbgPrintEx() which doesn’t make too much sense, IMO. WinDbg
> > won’t
> > > display info level prints by default and it is easy to forget them
> in
> > the
> > > code. DebugView displays them all which is probably because it
hooks
> > debug
> > > print API and not the routine which really prints.
> > >
> > > Next unwanted result of DbgPrintEx() is that incompetent support
> > persons
> > > don’t see debug prints in WinDbg and close reports about them as
> “not
> > > reproducible”. Even when log and information about using DebugView
> was
> >
> > > included. And DebugView is now MS tool
> > >
> > > An idea: what if DTM test warn about DbgPrint/Ex imports? It
checks
> > for
> > > unwanted imports anyway (at least HCT did).
> > >
> > > Best regards,
> > >
> > > Michal Vodicka
> > > UPEK, Inc.
> > > [xxxxx@upek.com, http://www.upek.com]
> > >
> > > >
> > > > ----------
> > > > From:
> > > >
> >
>
xxxxx@lists.osr.com[SMTP:xxxxx@lists.osr.com
> >]
> > > > on behalf of cristalink[SMTP:xxxxx@cristalink.com]
> > > > Reply To: Windows System Software Devs Interest List
> > > > Sent: Wednesday, December 06, 2006 8:14 PM
> > > > To: Windows System Software Devs Interest List
> > > > Subject: Re:[ntdev] Strange DebugOutStrung from cdrom driver
> > > >
> > > > Check classpnp and cdrom sources in DDK/src/storage.
> > > >
> > > >
> > > >
> > > > “Mark Shnaider” wrote in message
> news:xxxxx@ntdev…
> > > > Hello,
> > > > We have installed Vista (Release Build 6000) on 3 PC .On two
PC
> > > > DebugView
> > > > each second
> > > > get message " Will not retry; Sense/ASC/ASCQ of 02/3a/00" .
> > > > I found this text in the cdrom.sys driver . Why driver send
this
> > message.
> > > > How can I disable it?Stab.
> > > > Thanks in advice
> > > > Mark
> > > >
> > > >
> > > >
> > > >
> > > >
> > > >
> > > >
> > > >
> > > >
> > > >
> > > > —
> > > > Questions? First check the Kernel Driver FAQ at
> > > > http://www.osronline.com/article.cfm?id=256
> > > >
> > > > To unsubscribe, visit the List Server section of OSR Online at
> > > > http://www.osronline.com/page.cfm?name=ListServer
> > > >
> > >
> > >
> > >
> > > —
> > > Questions? First check the Kernel Driver FAQ at
> > http://www.osronline.com/article.cfm?id=256
> > >
> > > To unsubscribe, visit the List Server section of OSR Online at
> > http://www.osronline.com/page.cfm?name=ListServer
> > >
> >
> > —
> > Questions? First check the Kernel Driver FAQ at
> > http://www.osronline.com/article.cfm?id=256
> >
> > To unsubscribe, visit the List Server section of OSR Online at
> > http://www.osronline.com/page.cfm?name=ListServer
> >
> >
> > —
> > Questions? First check the Kernel Driver FAQ at
> http://www.osronline.com/article.cfm?id=256
> >
> > To unsubscribe, visit the List Server section of OSR Online at
> http://www.osronline.com/page.cfm?name=ListServer
> >
>
> —
> Questions? First check the Kernel Driver FAQ at>
> http://www.osronline.com/article.cfm?id=256
>
> To unsubscribe, visit the List Server section of OSR Online at
> http://www.osronline.com/page.cfm?name=ListServer
>
>
> —
> Questions? First check the Kernel Driver FAQ at
http://www.osronline.com/article.cfm?id=256
>
> To unsubscribe, visit the List Server section of OSR Online at
http://www.osronline.com/page.cfm?name=ListServer
>
Henry,
thanks for the offer but I prefer to sending flame mails to this list
BTW, why did you use debug prints at all when you have this nice and easy to use ETW?
You should also consider DebugView is probably much more widely used than kernel debugger. Everybody who uses it would see these strings at free builds and guess what they’d say about MS It is understandable how it could get into RTM only for developers who know how DbgPrintEx() works. For others, probably majority of DebugView users, it’d be just sign of MS carelessness.
Personaly, I prefer DebugView in the most cases. Traces from my drivers or apps can be captured by any user who encounters a problem. It is just matter to send them registry settings to enable traces, link to the web page and simple instructions. Asking them to use kernel debugger is quite unrealistic. Same with our QA.
I could request DebugView change but I’m not sure if it is desired. It already allows custom filtering and messages can be filtered according to their text (which is what I do with unwanted OS messages). Maybe if it is optional, settings can be changed on the fly and the default is to display all messages. If DebugView respects system wide settings, it wouldn’t display DbgPrint() output at Vista until enabled in the registry. Which would complicate instructions for users. So by default it should behave like now and the only gain of adding new filters would be better performance if unwanted strings are filtered by OS. Which is only considerable if some OS component barfs a lot of messages.
Currently, I still use DbgPrint() instead of DbgPrintEx(). If I ever change it, I’ll use DPFLTR_ERROR_LEVEL to eliminate Vista DbgPrint() mapping to Ex version. I have my own traces control which is more flexible than DbgPrintEx() allows and I send driver with debug prints to users only if really want to see traces according to my settings. I believe DbgPrint() mapping was Really Bad Idea™ which causes only trouble for us. I can understand adding DbgPrintEx() to make central settings for MS drivers but why to enforce others, who can’t add their IDs, to use it?
BTW, changing the code without changing all macro calls isn’t quite easy. My macros use “DbgPrint Params” construction where Params already contains parens. It can’t be changed to DbgPrintEx() without using compiler which allows variable macro parameters. The latest WDK fortunately contains such compiler but it wasn’t possible prior its release. Next reason why mapping is bad idea and I’d say it was rather arrogant decision.
Best regards,
Michal Vodicka
UPEK, Inc.
[xxxxx@upek.com, http://www.upek.com]
From: xxxxx@lists.osr.com[SMTP:xxxxx@lists.osr.com] on behalf of Henry Gabryjelski[SMTP:xxxxx@microsoft.com]
Reply To: Windows System Software Devs Interest List
Sent: Thursday, December 07, 2006 7:01 PM
To: Windows System Software Devs Interest List
Subject: RE:[ntdev] Strange DebugOutStrung from cdrom driverFirst off, the developer who made this tragic error was me. I did it.
I’m to blame. Sorry.Let’s scope the impact of this problem:
- Debugging FRE builds, this is not seen by default. (using kernel
debugger of course)- Even Debugging in CHK builds, it isn’t seen by default.
- It adds a few strings to the binary image; nothing sensitive.
- The only way to have it show up is to enable “DPFLTR_CDROM_ID” to
warning or higher level.You are correct that this should not have been checked in w/o the use of
macros. As a consolation, Michal, you can now send me up to one flame
mail a month for the next year.Since you know that we aren’t likely able to modify this, perhaps you
can request that DebugView be updated to respect the component ID and>
levels instead of printing everything? That would also make the utility
more useful on CHK builds….
-----Original Message-----
From: Michal Vodicka [mailto:xxxxx@upek.com]
Sent: Wednesday, December 06, 2006 2:49 PM
Subject: RE: Strange DebugOutStrung from cdrom driverI quite understand problem with resources and priorities, we have it,
too. In my experience the worst problem is when developer doesn’t even
know there is a problem and learn support persons if they can’t
reproduce a problem, it doesn’t mean it doesn’t exist. In this case it
is one minute fix for a tired developer who doesn’t dare to do anything
more complicated this day. In the beta phase and if it is logged into
bug tracking database even as the lowest priority problem. After release
everything is much harder.BTW, in such cases open source projects have a big advantage. A
volunteer makes a fix and developer or project manager only approves it.Best regards,
Michal Vodicka
UPEK, Inc.
[xxxxx@upek.com, http://www.upek.com]> ----------
> From:
xxxxx@lists.osr.com[SMTP:xxxxx@lists.osr.com
] on behalf of Peter Wieland[SMTP:xxxxx@windows.microsoft.com]
> Reply To: Windows System Software Devs Interest List
> Sent: Wednesday, December 06, 2006 11:21 PM
> To: Windows System Software Devs Interest List
> Subject: RE: [ntdev] Strange DebugOutStrung from cdrom driver
>
> In the grand scope of things it allows us to spend our resources (dev
&
> test) fixing big issues.
>
> However if it’s any consolation it also annoys us when we can’t fix
this
> piddly stuff too.
>
> -p
> -----Original Message-----
> From: xxxxx@lists.osr.com
> [mailto:xxxxx@lists.osr.com] On Behalf Of Michal Vodicka
> Sent: Wednesday, December 06, 2006 2:10 PM
> To: Windows System Software Devs Interest List
> Subject: RE: [ntdev] Strange DebugOutStrung from cdrom driver
>
> Thanks, at least this way works
> because reported several such problems during beta phase and was
> dissatisfied with reports handling.
>
> The next possibility is to fix it locally, build driver and replace
> installed version. Is the WDK 6000 version equivallent with driver
> supplied with Vista RTM?
>
> Best regards,
>
> Michal Vodicka
> UPEK, Inc.
> [xxxxx@upek.com, http://www.upek.com]
>
>
> > ----------
> > From:
>
xxxxx@lists.osr.com[SMTP:xxxxx@lists.osr.com
> ] on behalf of Peter Wieland[SMTP:xxxxx@windows.microsoft.com]
> > Reply To: Windows System Software Devs Interest List
> > Sent: Wednesday, December 06, 2006 10:36 PM
> > To: Windows System Software Devs Interest List
> > Subject: RE: [ntdev] Strange DebugOutStrung from cdrom driver
> >
> > I doubt it’s something we’ll issue a QFE for. However I’ve told the
> > storage team about it … they’ll add it to the list of things to
> > cleanup in whatever the next version of CDROM is.
> >
> > In the mean time they’ve suggested putting a CD in the drive, as
they
> > suspect the error is coming back because the drive is empty.
Putting
> > something in there should stop the errors and keep it from filling
up
> > your viewer.
> >
> > -p
> >
> > -----Original Message-----
> > From: xxxxx@lists.osr.com
> > [mailto:xxxxx@lists.osr.com] On Behalf Of Michal
Vodicka
> > Sent: Wednesday, December 06, 2006 12:16 PM
> > To: Windows System Software Devs Interest List
> > Subject: RE: [ntdev] Strange DebugOutStrung from cdrom driver
> >
> > I knew CDROM drivers were in DDK/WDK
> there
> > are up-to-date sources which contain this nonsense.
> >
> > OP can now report problem and point to exact lines of code which>
> should
> > be fixed. However, I don’t expect it’d persuade support persons
> problem
> > even exists…
> >
> > Best regards,
> >
> > Michal Vodicka
> > UPEK, Inc.
> > [xxxxx@upek.com, http://www.upek.com]
> >
> >
> > > ----------
> > > From:
> >
>
xxxxx@lists.osr.com[SMTP:xxxxx@lists.osr.com
> > ] on behalf of cristalink[SMTP:xxxxx@cristalink.com]
> > > Reply To: Windows System Software Devs Interest List>
> > > Sent: Wednesday, December 06, 2006 9:07 PM
> > > To: Windows System Software Devs Interest List
> > > Subject: Re:[ntdev] Strange DebugOutStrung from cdrom driver
> > >
> > > >>To my surprise WDK (both 5600 and 6000) contain sources with it.
> > >
> > > They have been there since at least NT4.
> > >
> > >
> > >
> > >
> > > “Michal Vodicka” wrote in message
> > > > news:xxxxx@ntdev…
> > > > Good idea. To my surprise WDK (both 5600 and 6000) contain sources
> > > with it.
> > > > src\storage\class\cdrom\sense.c calls DbgPrintEx() directly
> without
> > > #if DBG.
> > > > Shame on the developer who did it. Not because forgot it in the
> code
> > > which
> > > > is quite understandable but because called the function directly
> and
> > > not via
> > > > a macro. Which is the only way how to make sure release build
> > doesn’t
> > > > contains forgotten debug prints.
> > > >
> > > > It isn’t the first and the only case at Vista. I’m affraid it is
> > > result of
> > > > adding DbgPrintEx() which doesn’t make too much sense, IMO. WinDbg
> > > won’t
> > > > display info level prints by default and it is easy to forget them
> > in
> > > the
> > > > code. DebugView displays them all which is probably because it
> hooks
> > > debug
> > > > print API and not the routine which really prints.
> > > >
> > > > Next unwanted result of DbgPrintEx() is that incompetent support
> > > persons
> > > > don’t see debug prints in WinDbg and close reports about them as
> > “not
> > > > reproducible”. Even when log and information about using DebugView
> > was
> > >
> > > > included. And DebugView is now MS tool
> > > >
> > > > An idea: what if DTM test warn about DbgPrint/Ex imports? It
> checks
> > > for
> > > > unwanted imports anyway (at least HCT did).
> > > >
> > > > Best regards,
> > > >
> > > > Michal Vodicka
> > > > UPEK, Inc.
> > > > [xxxxx@upek.com, http://www.upek.com]
> > > >
> > > > >
> > > > > ----------
> > > > > From:
> > > > >
> > >
> >
> xxxxx@lists.osr.com[SMTP:xxxxx@lists.osr.com
> > >]
> > > > > on behalf of cristalink[SMTP:xxxxx@cristalink.com]
> > > > > Reply To: Windows System Software Devs Interest List
> > > > > Sent: Wednesday, December 06, 2006 8:14 PM
> > > > > To: Windows System Software Devs Interest List
> > > > > Subject: Re:[ntdev] Strange DebugOutStrung from cdrom driver
> > > > >
> > > > > Check classpnp and cdrom sources in DDK/src/storage.
> > > > >
> > > > >
> > > > >
> > > > > “Mark Shnaider” wrote in message
> > news:xxxxx@ntdev…
> > > > > Hello,
> > > > > We have installed Vista (Release Build 6000) on 3 PC .On two
> PC
> > > > > DebugView
> > > > > each second
> > > > > get message " Will not retry; Sense/ASC/ASCQ of 02/3a/00" .
> > > > > I found this text in the cdrom.sys driver . Why driver send
> this
> > > message.
> > > > > How can I disable it?Stab.
> > > > > Thanks in advice
> > > > > Mark
> > > > >
> > > > >
> > > > >
> > > > >
> > > > >
> > > > >
> > > > >
> > > > >
> > > > >
> > > > >
> > > > > —
> > > > > Questions? First check the Kernel Driver FAQ at
> > > > > http://www.osronline.com/article.cfm?id=256
> > > > >
> > > > > To unsubscribe, visit the List Server section of OSR Online at
> > > > > http://www.osronline.com/page.cfm?name=ListServer
> > > > >
> > > >
> > > >
> > > >
> > > > —
> > > > Questions? First check the Kernel Driver FAQ at>
> > > http://www.osronline.com/article.cfm?id=256
> > > >
> > > > To unsubscribe, visit the List Server section of OSR Online at
> > > http://www.osronline.com/page.cfm?name=ListServer
> > > >
> > >
> > > —
> > > Questions? First check the Kernel Driver FAQ at
> > > http://www.osronline.com/article.cfm?id=256
> > >
> > > To unsubscribe, visit the List Server section of OSR Online at
> > > http://www.osronline.com/page.cfm?name=ListServer
> > >
> > >
> > > —
> > > Questions? First check the Kernel Driver FAQ at
> > http://www.osronline.com/article.cfm?id=256
> > >
> > > To unsubscribe, visit the List Server section of OSR Online at
> > http://www.osronline.com/page.cfm?name=ListServer
> > >
> >
> > —
> > Questions? First check the Kernel Driver FAQ at>
> > http://www.osronline.com/article.cfm?id=256
> >
> > To unsubscribe, visit the List Server section of OSR Online at
> > http://www.osronline.com/page.cfm?name=ListServer
> >
> >
> > —
> > Questions? First check the Kernel Driver FAQ at
> http://www.osronline.com/article.cfm?id=256
> >
> > To unsubscribe, visit the List Server section of OSR Online at
> http://www.osronline.com/page.cfm?name=ListServer
> >
>
> —
> Questions? First check the Kernel Driver FAQ at http://www.osronline.com/article.cfm?id=256
>
> To unsubscribe, visit the List Server section of OSR Online at http://www.osronline.com/page.cfm?name=ListServer
>
That’s (“who doesn’t dare do anything more complicated this day”) a good
point. Everybody has days that are meant for addressing problems like
this.
mm
>> xxxxx@upek.com 2006-12-06 17:49 >>>
I quite understand problem with resources and priorities, we have it,
too. In my experience the worst problem is when developer doesn’t even
know there is a problem and learn support persons if they can’t
reproduce a problem, it doesn’t mean it doesn’t exist. In this case it
is one minute fix for a tired developer who doesn’t dare to do anything
more complicated this day. In the beta phase and if it is logged into
bug tracking database even as the lowest priority problem. After release
everything is much harder.
BTW, in such cases open source projects have a big advantage. A
volunteer makes a fix and developer or project manager only approves it.
Best regards,
Michal Vodicka
UPEK, Inc.
[xxxxx@upek.com, http://www.upek.com]
From:
xxxxx@lists.osr.com[SMTP:xxxxx@lists.osr.com]
on behalf of Peter Wieland[SMTP:xxxxx@windows.microsoft.com]
Reply To: Windows System Software Devs Interest List
Sent: Wednesday, December 06, 2006 11:21 PM
To: Windows System Software Devs Interest List
Subject: RE: [ntdev] Strange DebugOutStrung from cdrom driverIn the grand scope of things it allows us to spend our resources (dev
&
test) fixing big issues.However if it’s any consolation it also annoys us when we can’t fix
this
piddly stuff too.-p
-----Original Message-----
From: xxxxx@lists.osr.com
[mailto:xxxxx@lists.osr.com] On Behalf Of Michal
Vodicka
Sent: Wednesday, December 06, 2006 2:10 PM
To: Windows System Software Devs Interest List
Subject: RE: [ntdev] Strange DebugOutStrung from cdrom driverThanks, at least this way works
because reported several such problems during beta phase and was
dissatisfied with reports handling.The next possibility is to fix it locally, build driver and replace
installed version. Is the WDK 6000 version equivallent with driver
supplied with Vista RTM?Best regards,
Michal Vodicka
UPEK, Inc.
[xxxxx@upek.com, http://www.upek.com]> ----------
> From:
xxxxx@lists.osr.com[SMTP:xxxxx@lists.osr.com
] on behalf of Peter Wieland[SMTP:xxxxx@windows.microsoft.com]
> Reply To: Windows System Software Devs Interest List
> Sent: Wednesday, December 06, 2006 10:36 PM
> To: Windows System Software Devs Interest List
> Subject: RE: [ntdev] Strange DebugOutStrung from cdrom driver
>
> I doubt it’s something we’ll issue a QFE for. However I’ve told
the
> storage team about it … they’ll add it to the list of things to
> cleanup in whatever the next version of CDROM is.
>
> In the mean time they’ve suggested putting a CD in the drive, as
they
> suspect the error is coming back because the drive is empty.
Putting
> something in there should stop the errors and keep it from filling
up
> your viewer.
>
> -p
>
> -----Original Message-----
> From: xxxxx@lists.osr.com
> [mailto:xxxxx@lists.osr.com] On Behalf Of Michal
Vodicka
> Sent: Wednesday, December 06, 2006 12:16 PM
> To: Windows System Software Devs Interest List
> Subject: RE: [ntdev] Strange DebugOutStrung from cdrom driver
>
> I knew CDROM drivers were in DDK/WDK
there
> are up-to-date sources which contain this nonsense.
>
> OP can now report problem and point to exact lines of code which
should
> be fixed. However, I don’t expect it’d persuade support persons
problem
> even exists…
>
> Best regards,
>
> Michal Vodicka
> UPEK, Inc.
> [xxxxx@upek.com, http://www.upek.com]
>
>
> > ----------
> > From:
>
xxxxx@lists.osr.com[SMTP:xxxxx@lists.osr.com
> ] on behalf of cristalink[SMTP:xxxxx@cristalink.com]
> > Reply To: Windows System Software Devs Interest List>
> > Sent: Wednesday, December 06, 2006 9:07 PM
> > To: Windows System Software Devs Interest List
> > Subject: Re:[ntdev] Strange DebugOutStrung from cdrom
driver
> >
> > >>To my surprise WDK (both 5600 and 6000) contain sources with
it.
> >
> > They have been there since at least NT4.
> >
> >
> >
> >
> > “Michal Vodicka” wrote in message
> > > news:xxxxx@ntdev…
> > > Good idea. To my surprise WDK (both 5600 and 6000) contain
sources
> > with it.
> > > src\storage\class\cdrom\sense.c calls DbgPrintEx() directly
without
> > #if DBG.
> > > Shame on the developer who did it. Not because forgot it in the
code
> > which
> > > is quite understandable but because called the function directly
and
> > not via
> > > a macro. Which is the only way how to make sure release build
> doesn’t
> > > contains forgotten debug prints.
> > >
> > > It isn’t the first and the only case at Vista. I’m affraid it is
> > result of
> > > adding DbgPrintEx() which doesn’t make too much sense, IMO.
WinDbg
> > won’t
> > > display info level prints by default and it is easy to forget
them
> in
> > the
> > > code. DebugView displays them all which is probably because it
hooks
> > debug
> > > print API and not the routine which really prints.
> > >
> > > Next unwanted result of DbgPrintEx() is that incompetent support
> > persons
> > > don’t see debug prints in WinDbg and close reports about them as
> “not
> > > reproducible”. Even when log and information about using
DebugView
> was
> >
> > > included. And DebugView is now MS tool
> > >
> > > An idea: what if DTM test warn about DbgPrint/Ex imports? It
checks
> > for
> > > unwanted imports anyway (at least HCT did).
> > >
> > > Best regards,
> > >
> > > Michal Vodicka
> > > UPEK, Inc.
> > > [xxxxx@upek.com, http://www.upek.com]
> > >
> > > >
> > > > ----------
> > > > From:
> > > >
> >
>
xxxxx@lists.osr.com[SMTP:xxxxx@lists.osr.com
> >]
> > > > on behalf of cristalink[SMTP:xxxxx@cristalink.com]
> > > > Reply To: Windows System Software Devs Interest List
> > > > Sent: Wednesday, December 06, 2006 8:14 PM
> > > > To: Windows System Software Devs Interest List
> > > > Subject: Re:[ntdev] Strange DebugOutStrung from cdrom driver
> > > >
> > > > Check classpnp and cdrom sources in DDK/src/storage.
> > > >
> > > >
> > > >
> > > > “Mark Shnaider” wrote in message
> news:xxxxx@ntdev…
> > > > Hello,
> > > > We have installed Vista (Release Build 6000) on 3 PC .On two
PC
> > > > DebugView
> > > > each second
> > > > get message " Will not retry; Sense/ASC/ASCQ of 02/3a/00" .
> > > > I found this text in the cdrom.sys driver . Why driver send
this
> > message.
> > > > How can I disable it?Stab.
> > > > Thanks in advice
> > > > Mark
> > > >
> > > >
> > > >
> > > >
> > > >
> > > >
> > > >
> > > >
> > > >
> > > >
> > > > —
> > > > Questions? First check the Kernel Driver FAQ at
> > > > http://www.osronline.com/article.cfm?id=256
> > > >
> > > > To unsubscribe, visit the List Server section of OSR Online at
> > > > http://www.osronline.com/page.cfm?name=ListServer
> > > >
> > >
> > >
> > >
> > > —
> > > Questions? First check the Kernel Driver FAQ at
> > http://www.osronline.com/article.cfm?id=256
> > >
> > > To unsubscribe, visit the List Server section of OSR Online at
> > http://www.osronline.com/page.cfm?name=ListServer
> > >
> >
> > —
> > Questions? First check the Kernel Driver FAQ at
> > http://www.osronline.com/article.cfm?id=256
> >
> > To unsubscribe, visit the List Server section of OSR Online at
> > http://www.osronline.com/page.cfm?name=ListServer
> >
> >
> > —
> > Questions? First check the Kernel Driver FAQ at
> http://www.osronline.com/article.cfm?id=256
> >
> > To unsubscribe, visit the List Server section of OSR Online at
> http://www.osronline.com/page.cfm?name=ListServer
> >
>
> —
> Questions? First check the Kernel Driver FAQ at>
> http://www.osronline.com/article.cfm?id=256
>
> To unsubscribe, visit the List Server section of OSR Online at
> http://www.osronline.com/page.cfm?name=ListServer
>
>
> —
> Questions? First check the Kernel Driver FAQ at
http://www.osronline.com/article.cfm?id=256
>
> To unsubscribe, visit the List Server section of OSR Online at
http://www.osronline.com/page.cfm?name=ListServer
>
—
Questions? First check the Kernel Driver FAQ at
http://www.osronline.com/article.cfm?id=256
To unsubscribe, visit the List Server section of OSR Online at
http://www.osronline.com/page.cfm?name=ListServer
Yes, but I think you guys might have missed Peter Wieland’s point. Most devs WANT to fix such untidy code (which is all this really is, far from a tragedy) when they find it. But, unfortunately, by the time you’re into later beta releases and RCs you’re NOT ALLOWED to fix such things. They don’t meet the minimum level of “seriousness” for making a fix and checking it in late in a release cycle.
In reality, the only way low priority clean-up like this gets done is either (a) early in the development/test cycle when checkins are unrestricted, or (b) when a different bug of sufficiently high priority get fixed and the dev just “sneaks in” some general cleanup.
Henry’s probably too good to have any high priority bugs late in the testing cycle (right Henry?) and so… you get to live with a single stray DbgPrint until, oh, probably LH Server or later. I sincerely hope this bug is the most annoying thing you have to deal with in Vista.
Peter
OSR
Any dev caught doing (b) is usually going to get some deservedly negative feedback, as well. It is bad practice.
Too many times, this “cleanup” or “trivial change” introduces new bugs late in the cycle and they delay release, or even worse, escape detection until after release.
Or it breaks dependencies nobody even knew existed. As a contrived, and deliberately evil example- assume some idiot was actually relying on those messages in some critical test procedure in an entirely different team struggling to complete their work- fine they did the wrong thing, there were smarter ways for them to do it, and it was an accident waiting to happen- Henry still would have made a change that put a multibillion dollar project’s release date in jeopardy, and it wasn’t even an important change to make. Innocent as it may be, it won’t escape notice in that event. You can wind up getting executive level attention of the most undesirable sort…
Any change has risks, even if some of them are vanishingly negligible (I can’t recall fixing a comment typo ever causing trouble, but I can conceive of ways it could be).
I can point to a specific (fortunately, not very significant) instance of such a case (or something very close to it) occurring in our own product, and disclosed elsewhere on this site. One of our devs put some “test code” in a code sample which he was using as a test vehicle for a required change, and it didn’t work, and in fact it demonstrated incorrect usage. That sample was released with that code in it, and someone later tried to use it, and was understandably confused and perplexed by the results.
That incident was troubling to me, because we had been using a process of QA review of code changes prior to approving new code for check-in, in part to avoid this very problem. But it fell into disuse (there was a sense of stepping outside the role, infringing on code ownership, too much redundancy, among others). Don’t know if the escape happened because we quit doing it, or because review just plain doesn’t always catch things, for that matter- audit trail wasn’t good enough to tell, and the problem wasn’t world-shattering enough to be worth any more post-mortem effort. What I saw was a process that failed to work, though…
I’d suggest that because of the above, this may actually be the worst time a problem like this get addressed. If you can’t think straight, go home, relax with family, friends, pets, whatever. Decompress and come back when you can do something useful. Better for everyone in the long run.
Yes, it is annoying (sometimes extremely so) to have to work in this manner. But so far, that’s the way it is.
We aren’t in disagreement; we have to deal with such things, too, and I understand things get worse with company size. But my point was a bit different. I complained about the way how support handles such problems. If they logged such case as the lowest priority problem and it won’t get fixed for reasons below, it’d be quite correct. But I reported something similar (not exactly messages discussed here) and they refused it as not-reproducible although had logs and name of driver binary which clearly contained these messages. That’s just stupid. It was is early development phase and took several months and repeated reports until it was logged and finally fixed.
I’d suggest that because of the above, this may actually be the worst time a problem like this get addressed. If you can’t think straight, go home, relax with family, friends, pets, whatever. Decompress and come back when you can do something useful. Better for everyone in the long run.
This way I’d stay at home most of the time
Best regards,
Michal Vodicka
UPEK, Inc.
[xxxxx@upek.com, http://www.upek.com]
From: xxxxx@lists.osr.com[SMTP:xxxxx@lists.osr.com] on behalf of Bob Kjelgaard[SMTP:xxxxx@microsoft.com]
Reply To: Windows System Software Devs Interest List
Sent: Friday, December 08, 2006 4:59 PM
To: Windows System Software Devs Interest List
Subject: RE: RE:[ntdev] Strange DebugOutStrung from cdrom driverAny dev caught doing (b) is usually going to get some deservedly negative feedback, as well. It is bad practice.
Too many times, this “cleanup” or “trivial change” introduces new bugs late in the cycle and they delay release, or even worse, escape detection until after release.
Or it breaks dependencies nobody even knew existed. As a contrived, and deliberately evil example- assume some idiot was actually relying on those messages in some critical test procedure in an entirely different team struggling to complete their work- fine they did the wrong thing, there were smarter ways for them to do it, and it was an accident waiting to happen- Henry still would have made a change that put a multibillion dollar project’s release date in jeopardy, and it wasn’t even an important change to make. Innocent as it may be, it won’t escape notice in that event. You can wind up getting executive level attention of the most undesirable sort…
Any change has risks, even if some of them are vanishingly negligible (I can’t recall fixing a comment typo ever causing trouble, but I can conceive of ways it could be).
I can point to a specific (fortunately, not very significant) instance of such a case (or something very close to it) occurring in our own product, and disclosed elsewhere on this site. One of our devs put some “test code” in a code sample which he was using as a test vehicle for a required change, and it didn’t work, and in fact it demonstrated incorrect usage. That sample was released with that code in it, and someone later tried to use it, and was understandably confused and perplexed by the results.
That incident was troubling to me, because we had been using a process of QA review of code changes prior to approving new code for check-in, in part to avoid this very problem. But it fell into disuse (there was a sense of stepping outside the role, infringing on code ownership, too much redundancy, among others). Don’t know if the escape happened because we quit doing it, or because review just plain doesn’t always catch things, for that matter- au> dit trail wasn’t good enough to tell, and the problem wasn’t world-shattering enough to be worth any more post-mortem effort. What I saw was a process that failed to work, though…
I’d suggest that because of the above, this may actually be the worst time a problem like this get addressed. If you can’t think straight, go home, relax with family, friends, pets, whatever. Decompress and come back when you can do something useful. Better for everyone in the long run.
Yes, it is annoying (sometimes extremely so) to have to work in this manner. But so far, that’s the way it is.
Questions? First check the Kernel Driver FAQ at http://www.osronline.com/article.cfm?id=256
To unsubscribe, visit the List Server section of OSR Online at http://www.osronline.com/page.cfm?name=ListServer