Hi,
just had some trouble with fabsf(). I didn't put #include on
top of the file. Unfortunately the code produced random results.
Wouldn't it be nice to get an error or even a warning in this case?
I tried the parser options. But ether it doesn't give me a warning in my
case, or it generates errors in the CSL include files.
Any ideas?
TIA Gustl
P.S.: no money for CCS 3.*
Missing Error/Warning CCS2.2
Started by ●November 6, 2008
Reply by ●November 6, 20082008-11-06
I'm fairly sure you can turn on warnings for missing function
declarations.
----- Original Message ----
From: Bernhard 'Gustl' Bauer
To: C6x
Sent: Thursday, November 6, 2008 12:00:48 PM
Subject: [c6x] Missing Error/Warning CCS2.2
Hi,
just had some trouble with fabsf(). I didn't put #include on
top of the file. Unfortunately the code produced random results.
Wouldn't it be nice to get an error or even a warning in this case?
I tried the parser options. But ether it doesn't give me a warning in my
case, or it generates errors in the CSL include files.
Any ideas?
TIA Gustl
P.S.: no money for CCS 3.*
----- Original Message ----
From: Bernhard 'Gustl' Bauer
To: C6x
Sent: Thursday, November 6, 2008 12:00:48 PM
Subject: [c6x] Missing Error/Warning CCS2.2
Hi,
just had some trouble with fabsf(). I didn't put #include on
top of the file. Unfortunately the code produced random results.
Wouldn't it be nice to get an error or even a warning in this case?
I tried the parser options. But ether it doesn't give me a warning in my
case, or it generates errors in the CSL include files.
Any ideas?
TIA Gustl
P.S.: no money for CCS 3.*
Reply by ●November 6, 20082008-11-06
try -pd option. I think that upgrade warnings as errors.
--Bhooshan
-----Original Message-----
From: Andrew Elder
Sent: 06/11/2008 10:49:13 pm
To: Bernhard 'Gustl' Bauer; C6x
Subject: Re: [c6x] Missing Error/Warning CCS2.2
I'm fairly sure you can turn on warnings for missing function declarations.
----- Original Message ----
From: Bernhard 'Gustl' Bauer
To: C6x
Sent: Thursday, November 6, 2008 12:00:48 PM
Subject: [c6x] Missing Error/Warning CCS2.2
Hi,
just had some trouble with fabsf(). I didn't put #include on
top of the file. Unfortunately the code produced random results.
Wouldn't it be nice to get an error or even a warning in this case?
I tried the parser options. But ether it doesn't give me a warning in my
case, or it generates errors in the CSL include files.
Any ideas?
TIA Gustl
P.S.: no money for CCS 3.*
--Bhooshan
-----Original Message-----
From: Andrew Elder
Sent: 06/11/2008 10:49:13 pm
To: Bernhard 'Gustl' Bauer; C6x
Subject: Re: [c6x] Missing Error/Warning CCS2.2
I'm fairly sure you can turn on warnings for missing function declarations.
----- Original Message ----
From: Bernhard 'Gustl' Bauer
To: C6x
Sent: Thursday, November 6, 2008 12:00:48 PM
Subject: [c6x] Missing Error/Warning CCS2.2
Hi,
just had some trouble with fabsf(). I didn't put #include on
top of the file. Unfortunately the code produced random results.
Wouldn't it be nice to get an error or even a warning in this case?
I tried the parser options. But ether it doesn't give me a warning in my
case, or it generates errors in the CSL include files.
Any ideas?
TIA Gustl
P.S.: no money for CCS 3.*
Reply by ●November 7, 20082008-11-07
Bauer,
There should have been several errors during the link cycle about undefined symbols.
If no such errors were displayed, then 'somewhere' those function names are
defined.
Find out where they are defined and hide them from the linker.
A expect your compile and link options includes the 'verbose' and similar options.
R. Williams
---------- Original Message -----------
From: "Bernhard 'Gustl' Bauer"
To: C6x
Sent: Thu, 06 Nov 2008 18:00:48 +0100
Subject: [c6x] Missing Error/Warning CCS2.2
> Hi,
>
> just had some trouble with fabsf(). I didn't put #include on
> top of the file. Unfortunately the code produced random results.
>
> Wouldn't it be nice to get an error or even a warning in this case?
>
> I tried the parser options. But ether it doesn't give me a warning in
> my case, or it generates errors in the CSL include files.
>
> Any ideas?
>
> TIA Gustl
>
> P.S.: no money for CCS 3.*
>
>
There should have been several errors during the link cycle about undefined symbols.
If no such errors were displayed, then 'somewhere' those function names are
defined.
Find out where they are defined and hide them from the linker.
A expect your compile and link options includes the 'verbose' and similar options.
R. Williams
---------- Original Message -----------
From: "Bernhard 'Gustl' Bauer"
To: C6x
Sent: Thu, 06 Nov 2008 18:00:48 +0100
Subject: [c6x] Missing Error/Warning CCS2.2
> Hi,
>
> just had some trouble with fabsf(). I didn't put #include on
> top of the file. Unfortunately the code produced random results.
>
> Wouldn't it be nice to get an error or even a warning in this case?
>
> I tried the parser options. But ether it doesn't give me a warning in
> my case, or it generates errors in the CSL include files.
>
> Any ideas?
>
> TIA Gustl
>
> P.S.: no money for CCS 3.*
>
>
Reply by ●November 7, 20082008-11-07
Andrew Elder schrieb:
> I'm fairly sure you can turn on warnings for missing function declarations.
>
I have been looking for this. I didn't find anything. Does anybody know
how I can do this?
Gustl
> I'm fairly sure you can turn on warnings for missing function declarations.
>
I have been looking for this. I didn't find anything. Does anybody know
how I can do this?
Gustl
Reply by ●November 7, 20082008-11-07
Richard Williams schrieb:
> Bauer,
Williams,
> There should have been several errors during the link cycle about
> undefined symbols.
No errors or warning about fabsf(). I would expect the compiler, not the
linker to produce such an error or warning.
> If no such errors were displayed, then 'somewhere' those function names are
> defined.
> Find out where they are defined and hide them from the linker.
I did a search on all files for fabsf. I only found the lines where I
use fabsf.
> A expect your compile and link options includes the 'verbose' and
> similar options.
That doesn't change a thing.
Gustl
> Bauer,
Williams,
> There should have been several errors during the link cycle about
> undefined symbols.
No errors or warning about fabsf(). I would expect the compiler, not the
linker to produce such an error or warning.
> If no such errors were displayed, then 'somewhere' those function names are
> defined.
> Find out where they are defined and hide them from the linker.
I did a search on all files for fabsf. I only found the lines where I
use fabsf.
> A expect your compile and link options includes the 'verbose' and
> similar options.
That doesn't change a thing.
Gustl
Reply by ●November 7, 20082008-11-07
Bhooshan Iyer schrieb:
> try -pd option. I think that upgrade warnings as errors.
>
> --Bhooshan
I don't get a warning in the first place!
I tried several -pd options - witout success.
> try -pd option. I think that upgrade warnings as errors.
>
> --Bhooshan
I don't get a warning in the first place!
I tried several -pd options - witout success.
Reply by ●November 7, 20082008-11-07
Gust1,
The compiler does not list missing symbol information, unless a symbol is
referenced that is not defined in any manner within the scope of that file.
So if the library header file is included, then the symbol will not be listed as
an error/warning
It is the linker, which is trying to gather all the object code into a single
file (.out) that will complain if the code for a specific called symbol/function
is not found.
R. Williams
---------- Original Message -----------
From: "Bernhard 'Gustl' Bauer"
To: Richard Williams
Cc: C6x
Sent: Fri, 07 Nov 2008 08:12:24 +0100
Subject: Re: [c6x] Missing Error/Warning CCS2.2
> Richard Williams schrieb:
> > Bauer,
>
> Williams,
>
> > There should have been several errors during the link cycle about
> > undefined symbols.
>
> No errors or warning about fabsf(). I would expect the compiler, not
> the linker to produce such an error or warning.
>
> > If no such errors were displayed, then 'somewhere' those function names are
> > defined.
> > Find out where they are defined and hide them from the linker.
>
> I did a search on all files for fabsf. I only found the lines where I
> use fabsf.
>
> > A expect your compile and link options includes the 'verbose' and
> > similar options.
>
> That doesn't change a thing.
>
> Gustl
------- End of Original Message -------
The compiler does not list missing symbol information, unless a symbol is
referenced that is not defined in any manner within the scope of that file.
So if the library header file is included, then the symbol will not be listed as
an error/warning
It is the linker, which is trying to gather all the object code into a single
file (.out) that will complain if the code for a specific called symbol/function
is not found.
R. Williams
---------- Original Message -----------
From: "Bernhard 'Gustl' Bauer"
To: Richard Williams
Cc: C6x
Sent: Fri, 07 Nov 2008 08:12:24 +0100
Subject: Re: [c6x] Missing Error/Warning CCS2.2
> Richard Williams schrieb:
> > Bauer,
>
> Williams,
>
> > There should have been several errors during the link cycle about
> > undefined symbols.
>
> No errors or warning about fabsf(). I would expect the compiler, not
> the linker to produce such an error or warning.
>
> > If no such errors were displayed, then 'somewhere' those function names are
> > defined.
> > Find out where they are defined and hide them from the linker.
>
> I did a search on all files for fabsf. I only found the lines where I
> use fabsf.
>
> > A expect your compile and link options includes the 'verbose' and
> > similar options.
>
> That doesn't change a thing.
>
> Gustl
------- End of Original Message -------
Reply by ●November 7, 20082008-11-07
Gust1,
I have CCS 3.1 at the moment, not CCS 2.2.
For CCS 3.1, the compiler/linker parameters of concern are:
--verbose
-on2
-o3
-os
-b
-pdf
-pden
-pdr
-pdv
-w
-x
This will result in (possibly) a .err file for each source file
This will give maximum verbiage in the 'output' window during compile and link
operations.
R. Williams
---------- Original Message -----------
From: "Bernhard 'Gustl' Bauer"
To: Andrew Elder
Cc: C6x
Sent: Fri, 07 Nov 2008 08:03:11 +0100
Subject: Re: [c6x] Missing Error/Warning CCS2.2
> Andrew Elder schrieb:
> >
> >
> > I'm fairly sure you can turn on warnings for missing function declarations.
> > I have been looking for this. I didn't find anything. Does anybody
> know how I can do this?
>
> Gustl
------- End of Original Message -------
I have CCS 3.1 at the moment, not CCS 2.2.
For CCS 3.1, the compiler/linker parameters of concern are:
--verbose
-on2
-o3
-os
-b
-pden
-pdr
-pdv
-w
-x
This will result in (possibly) a .err file for each source file
This will give maximum verbiage in the 'output' window during compile and link
operations.
R. Williams
---------- Original Message -----------
From: "Bernhard 'Gustl' Bauer"
To: Andrew Elder
Cc: C6x
Sent: Fri, 07 Nov 2008 08:03:11 +0100
Subject: Re: [c6x] Missing Error/Warning CCS2.2
> Andrew Elder schrieb:
> >
> >
> > I'm fairly sure you can turn on warnings for missing function declarations.
> > I have been looking for this. I didn't find anything. Does anybody
> know how I can do this?
>
> Gustl
------- End of Original Message -------
Reply by ●November 7, 20082008-11-07
Hi Bernhard,
Oh my goodness, how I do understand you - the compiler has swirled to
have too many options!
Regarding undeclared function prototypes I guess I already contacted TI
(and it seem that without success) on that: the compiler generates a
_remark_ for that situation, not a warning, in spite of C that _requires_
every object _has_ to be declared first then used. Remarks are off by
default, to turn them on use:
>From the command prompt use these options:
-pdv verbose diagnostics
-pdr enable remarks
-pdf stderr to a file
In CCS environment go to:
Project->Build Options->Compiler->Diagnostics, in there mark "Issue Nonserious
Warnings (-pdr)" checkbox.
That stil makes me wander - calling an undeclared may result in a wrong program
behavior, but then why on Earth TI considers them as "nonserious warning"???
Rgds,
Andrew
> Subject Missing Error/Warning CCS2.2
> Posted by: "Bernhard 'Gustl' Bauer" gustl@
> Date: Thu Nov 6, 2008 9:01 am ((PST))
>
> Hi,
>
> just had some trouble with fabsf(). I didn't put #include on
> top of the file. Unfortunately the code produced random results.
>
> Wouldn't it be nice to get an error or even a warning in this case?
>
> I tried the parser options. But ether it doesn't give me a warning in my
> case, or it generates errors in the CSL include files.
>
> Any ideas?
>
> TIA Gustl
>
> P.S.: no money for CCS 3.*
>
Oh my goodness, how I do understand you - the compiler has swirled to
have too many options!
Regarding undeclared function prototypes I guess I already contacted TI
(and it seem that without success) on that: the compiler generates a
_remark_ for that situation, not a warning, in spite of C that _requires_
every object _has_ to be declared first then used. Remarks are off by
default, to turn them on use:
>From the command prompt use these options:
-pdv verbose diagnostics
-pdr enable remarks
-pdf stderr to a file
In CCS environment go to:
Project->Build Options->Compiler->Diagnostics, in there mark "Issue Nonserious
Warnings (-pdr)" checkbox.
That stil makes me wander - calling an undeclared may result in a wrong program
behavior, but then why on Earth TI considers them as "nonserious warning"???
Rgds,
Andrew
> Subject Missing Error/Warning CCS2.2
> Posted by: "Bernhard 'Gustl' Bauer" gustl@
> Date: Thu Nov 6, 2008 9:01 am ((PST))
>
> Hi,
>
> just had some trouble with fabsf(). I didn't put #include on
> top of the file. Unfortunately the code produced random results.
>
> Wouldn't it be nice to get an error or even a warning in this case?
>
> I tried the parser options. But ether it doesn't give me a warning in my
> case, or it generates errors in the CSL include files.
>
> Any ideas?
>
> TIA Gustl
>
> P.S.: no money for CCS 3.*
>
