Throw Exception and using Dispose?

[=?Utf-8?B?TVIgRQ==?=]

This may seem like a stupid question but in C#:

Say for instance I have a set of SQL processes that I run via
ExecuteReader() . These processes return several pieces of information to the
DataReader, including any SQL exception that occurred (This may include
messages I generated b/c something didn’t check out based on programmed
requirements, but its not a runtime exception as far as SQL or C# is
concerned).

I then check this exception (or programmed error message) and perform a TROW
NEW EXCEPTION() with a more friendlier message and perform some other tasks.
I realize throw exception is costly and only do it when you have to, but its
a great way to quickly access the Application_Err or() in global asx for
logging errors and generating notifications to the programmer. The problem is
I still need to go back and dispose of the class objects that where created
before the exception is thrown. If I try to dispose of the class first then I
get an Exception of the Exception. How can manually throw exceptions and
still go back to cleanup my code when Im not using a TRY/FINALLY b/c again
these are exception I threw, not the CLR?
Of course any code I place after throw new exception() gives me an
unreachable code detected during compile.

I know Im missing something really simple here

[Scott M.]

Why can't you instantiate your new exception and configure it with any
information you want, then dispose what you need to and then throw your
exception?
"MR E" <MR E@discussions.m icrosoft.comwrote in message
news:75******** *************** ***********@mic rosoft.com...

This may seem like a stupid question but in C#:

Say for instance I have a set of SQL processes that I run via
ExecuteReader() . These processes return several pieces of information to
the
DataReader, including any SQL exception that occurred (This may include
messages I generated b/c something didn't check out based on programmed
requirements, but its not a runtime exception as far as SQL or C# is
concerned).

I then check this exception (or programmed error message) and perform a
TROW
NEW EXCEPTION() with a more friendlier message and perform some other
tasks.
I realize throw exception is costly and only do it when you have to, but
its
a great way to quickly access the Application_Err or() in global asx for
logging errors and generating notifications to the programmer. The problem
is
I still need to go back and dispose of the class objects that where
created
before the exception is thrown. If I try to dispose of the class first
then I
get an Exception of the Exception. How can manually throw exceptions and
still go back to cleanup my code when Im not using a TRY/FINALLY b/c again
these are exception I threw, not the CLR?
Of course any code I place after throw new exception() gives me an
unreachable code detected during compile.

I know Im missing something really simple here