INTERNAL COMPILER ERROR C1001: msc1.cpp (line 1794) error at every std include file: stdio.h, window

[paul calvert]

I hope somewhere here has encountered and solved a similar problem in
the past.

1) on a new Win2000 PC:

installed Visual C++ 6.0
download & install single file Service Pack 5.0

2) try to build my gui and dll projects, whose project, workspace,
source files all resided on network drive mapped to H. The H mapping,
service packs, etc are all identical to a lab setup. The lab setup
builds no problems (Debug target). Lab setup is on Win NT 4.00.1381.
Build on Win2000 machine gave:

"INTERNAL COMPILER ERROR C1001: msc1.cpp (line 1794) error" at every
std include file: stdio.h, windows.h, ansi_c.h, etc

Went to a 3rd Pc--another NT4.0 The INTERNAL COMPILER ERROR C1001:
msc1.cpp (line 1794) error did not appear. Hmmm---is it a Win2000
problem?

3) ran Windows update on Win2000 machine: installed many patches (10
hrs worth) and the latest service pack 4.

4) after the OS updates, the gui then would build--just like the lab
NT machine. (debug mode) The INTERNAL COMPILER ERROR C1001: msc1.cpp
(line 1794) error disappeared. Pure magic.

5) started monkeying with project settings. By happen-stance, I
changed from Debug to Release configuration. After which, the internal
Compiler errors disappeared and then started getting linker errors on
some 3rd party lib calls

7) I copied the debug settings to the release settings manually. This
included:

under the link tab: change Input-> to ignore "libc". I don't know
what/why this was even in debug settings, but without it I got errors.
under the link tab: change General-> cut/pasted the 3rd party
libraries string from debug configuration
under the link tab: change General-> checked the 'Generate debug',
'Link incrementally', 'Generate mapfile' options

As best I can determine now the debug/release configuration options
are the same. Rebuilt with release mode set to active: Seems to build
OK now. NOTE: I have NO idea if the actual generated code actually
runs yet because of required target hardware availability (Hope to
find out soon). HOWEVER, even though the debug/release modes are now
seemingly identical in options, the compiler still pukes out the
"INTERNAL COMPILER ERROR C1001: msc1.cpp (line 1794) error" when the
project set configuration is "Debug". What can I conclude?

1) there seems to be several OS related issued with Win2000 and Visual
C++ 6.0 SP5. Perhaps someone else has already encountered/solved this
problem??

2) these bugs may be tied to the project options and Debug/Release
settings

3) the "INTERNAL COMPILER ERROR C1001: msc1.cpp (line 1794)" error is
apparently a frequent occurrence which Microsoft claims to have fixed
in the .net version. Note that Microsoft's bug tracking, points to
several possible causes for the error--none of which applied in my
situation. I found hundreds of website references of others having the
same problem and they had varying degrees of luck in solving it.
Unfortunately, most of these did not make any reference to the OS they
were running. Several folks indicated they switched to .net or Borland
because the problem couldn't be fixed and/or Microsoft refused to fix
with a service pack. Note that I am not using or generating any
pre-compiled headers-- the supposed usual cause of such an error.

4)I saw some references that indicated this type problem was
eliminated on 7.0--but not from Microsoft. Their website only
explicitly stated this for the .net version. Is this accurate?

This is exceedingly frustrating and I still have no idea about the
real cause or a real solution to the problem. I have merely stumbled
upon a hopefully functional work-around to the problem. I still want
to try the build on a fresh NT machine and see if any problems are
encountered at all.

If you have any light to shed... please.

Thanks,

Paul

 

[WW]

I hope somewhere here has encountered and solved a similar problem in
the past.

1) on a new Win2000 PC:

installed Visual C++ 6.0
download & install single file Service Pack 5.0

2) try to build my gui and dll projects, whose project, workspace,
source files all resided on network drive mapped to H. The H mapping,
service packs, etc are all identical to a lab setup. The lab setup
builds no problems (Debug target). Lab setup is on Win NT 4.00.1381.
Build on Win2000 machine gave:

"INTERNAL COMPILER ERROR C1001: msc1.cpp (line 1794) error" at every
std include file: stdio.h, windows.h, ansi_c.h, etc

http://www.parashift.com/c++-faq-lite/how-to-post.html#faq-5.9

 

[bo]

I'm not sure exactly what the problem with my org post was, but I
re-posted to comp.os.ms-windows.tools and .tools.misc. Sorry if my
org post ruffled your feathers.

Serbus,

Paul

 

[Ron Ruble]

I'm not sure exactly what the problem with my org post was, but I
re-posted to comp.os.ms-windows.tools and .tools.misc. Sorry if my
org post ruffled your feathers.

No ruffled feathers. The problem is your question had
nothing to do with the C++ -language-. It had to do with
a very compiler-specific error, and compiler-specific
behavior is not what this groups is about.

From the FAQ link referenced:

"Only post to comp.lang.c++ if your question is
about the C++ language itself"

Your question was not about the language. No harm
done; just something to be aware of for future questions.

 

[WW]

I'm not sure exactly what the problem with my org post was, but I
re-posted to comp.os.ms-windows.tools and .tools.misc. Sorry if my
org post ruffled your feathers.

Top-posting, rearranged.

http://www.parashift.com/c++-faq-lite/how-to-post.html#faq-5.4

I really wonder that from that one line (pointing you to a place which tells
you which is the right newsgroup to post your question to) how was it
possible to deduce that feathers has been ruffled ... or anything more than
the fact that your post had nothing to do with the C++ language so you have
been redirected. I am sorry, but I am really too dumb to follow that
thinking.

 

[bo]

I did not deduce that feathers were ruffled--I was just trying to be
nice by saying "IF" . In any case, after posting to the groups
suggested--no response there either. I guess I had false hope that a
more heavily used newsgroup would be more likely to cross paths with
someone encountering the problem even though the topic was not a C++
question per se. Is there a better group to post the question to?

Thanks,

Paul

 

[WW]

I did not deduce that feathers were ruffled--I was just trying to be
nice by saying "IF" . In any case, after posting to the groups
suggested--no response there either. I guess I had false hope that a
more heavily used newsgroup would be more likely to cross paths with
someone encountering the problem even though the topic was not a C++
question per se. Is there a better group to post the question to?

The billion Microsoft newsgroups on the Microsoft server.