Compiler "Warnings"

Jeremy G · 04-17-2005, 07:46 PM

Seriously whats up with compiler warnings? Warnings don't mean program failure, speciffically in this case my program works exactly as intended - beautifully I might add, but compiles with a healthy list of warnings.

As far as I've been concerned I've always considered compile warnings to be like "Hey, if you don't know what you're doing this could **** you up.." Of course I've always known what I was doing so I found them to be rediculous nagging.

So I am asking you, if you compile your programs, works exactly as intended, passes test cases out the wazoo, do you continue to modify the code eliminating "Warnings" ?

If you are curious, this is the list of warnings from my latest program. I'm not asking for help, I'm just asking for personal opinions.

Quote:

C:\MINGW32\bin>gcc -o main.c main.exe
gcc: main.exe: No such file or directory
gcc: No input files

C:\MINGW32\bin>gcc -o main.exe main.c
main.c:45: warning: data definition has no type or storage class
main.c: In function `bTreeInit':
main.c:108: warning: assignment from incompatible pointer type
main.c:109: warning: assignment from incompatible pointer type
main.c: In function `bTreeAdd':
main.c:129: warning: passing arg 1 of `bTreeAdd' from incompatible pointer type
main.c:131: warning: passing arg 1 of `bTreeAdd' from incompatible pointer type
main.c: In function `bTreeSearch':
main.c:146: warning: passing arg 1 of `strlwr' discards `const' from pointer tar
get type
main.c:146: warning: passing arg 1 of `strcpy' discards `const' from pointer tar
get type
main.c:152: warning: passing arg 1 of `bTreeSearch' from incompatible pointer ty
pe
main.c:154: warning: passing arg 1 of `bTreeSearch' from incompatible pointer ty
pe
main.c: In function `bTreePreOrder':
main.c:162: warning: passing arg 1 of `bTreePreOrder' from incompatible pointer
type
main.c:163: warning: passing arg 1 of `bTreePreOrder' from incompatible pointer
type
main.c: In function `bTreeDepth':
main.c:174: warning: passing arg 1 of `bTreeDepth' from incompatible pointer typ
e
main.c:175: warning: passing arg 1 of `bTreeDepth' from incompatible pointer typ
e
main.c: In function `bTreeWordCount':
main.c:191: warning: passing arg 1 of `bTreeWordCount' from incompatible pointer
type
main.c:192: warning: passing arg 1 of `bTreeWordCount' from incompatible pointer
type
main.c: In function `bTreePreOrderPop':
main.c:214: warning: return from incompatible pointer type
main.c:216: warning: passing arg 1 of `bTreePreOrderPop' from incompatible point
er type
main.c:218: warning: assignment from incompatible pointer type
main.c:219: warning: return from incompatible pointer type
main.c:219: warning: function returns address of local variable
main.c:222: warning: passing arg 1 of `bTreePreOrderPop' from incompatible point
er type
main.c:224: warning: assignment from incompatible pointer type
main.c:225: warning: return from incompatible pointer type
main.c:225: warning: function returns address of local variable
main.c:231: warning: return from incompatible pointer type
main.c:231: warning: function returns address of local variable
main.c: In function `sDocumentAdd':
main.c:247: warning: passing arg 1 of `sDocumentAdd' from incompatible pointer t
ype
main.c: In function `sDocumentInit':
main.c:251: warning: assignment from incompatible pointer type
main.c: In function `sDocumentPrint':
main.c:272: warning: passing arg 1 of `sDocumentPrint' from incompatible pointer
type
main.c: In function `queryFile':
main.c:411: warning: assignment from incompatible pointer type

B0bDole · 04-17-2005, 08:06 PM

those warnings could cause memory problems... leaks, miss-assignments, etc. Something you probably dont see while doing tests on the program.

gcn_zelda · 04-17-2005, 08:33 PM

Yeah. Those kinds of warnings make programs like Windows come to pass.

[/WindowsBash]

Jeremy G · 04-17-2005, 08:52 PM

Both of you were probably right.

Those warnings were actually pretty serious. And I've completely eliminated them.

How ever, some warnings are just superflous.

major_small · 04-18-2005, 01:00 AM

not really... the way I see it, you should try to minimize warnings by as much as possible... when I'm getting a warning, I find out what it means and if/how I can get rid of it. the standards were drawn up for a reason--there are reasons for every one of the warnings you get. it's not like they just said "let's throw this warning in there just for kicks".

I agree--you can compile and run a program with no problems if you don't explicitly cast a float to an integer, but that doesn't help the readability of the code, and it may actually remind you that you're going to be losing data there...

**nvoigt** · 04-18-2005, 02:18 AM

Quote:

So I am asking you, if you compile your programs, works exactly as intended, passes test cases out the wazoo, do you continue to modify the code eliminating "Warnings" ?

None of our programs at work is considered complete if there are still warnings in the release build.

**Salem** · 04-18-2005, 10:56 AM

> gcc -o main.c main.exe
OMG!!!
So you want to OVERWRITE your source file with the compiled code?
Be more careful with your commands man, otherwise you'll be in the same position as another poor sap who managed to trash all his code a couple of hours before the "deadline".

It's a good job the command line was invalid as well, otherwise you'd be scrabbling round for a backup (you do have backups right?)

skorman00 · 04-18-2005, 12:50 PM

The superflous warnings you're talking about are probably like "unsigned being compared against signed value" or "float to int conversion, possible loss of data." For most pieces of data they won't matter, but they could come to bite you in the arse.

joshdick · 04-18-2005, 01:01 PM

The most superfluous warnings I've received were from Visual Studio 6.0 telling me that I had identifiers that were too long. I wasn't personally using 300-character identifiers. The problem is that a vector of strings expands out to an identifier of obscene length explictly stating the type of character set being used and the allocator that the vector is using and so on. Eleven warnings every compile like that, but not a single problem caused by it in testing, so meh.

By the way, has anyone else ever run into that problem?

skorman00 · 04-18-2005, 01:15 PM

I know a buddy who swears by long and descriptive function/variable names. He'd come up with something like

Code:

TakeTheXValueAndCalculateTheCorrectYValue(int thisIsTheXThatIsUsedToComputeTheY)

<dramatization>
I'm going to poke fun at him now with this, thanks joshdick.

Some of you fellas/ladies may have already seen this, but it's always worth a second look.
http://www.effect.net.au/lukastan/hu...ssages-Mac.htm

pianorain · 04-18-2005, 01:26 PM

Quote:

Originally Posted by joshdick

By the way, has anyone else ever run into that problem?

Yup, I hated it. It could get pretty bad, especially if you nested your vector of strings even further into STL containers. VS.Net is much nicer about that.

The warning I'm working on eliminating right now at work involves the STL containers not having dll-interfaces. I don't think it's a problem, but it'd be nice not have 22 warnings every build.

joshdick · 04-18-2005, 01:36 PM

Quote:

Originally Posted by pianorain

I don't think it's a problem, but it'd be nice not have 22 warnings every build.

Yeah, that's how I felt. I knew there wasn't another truly wrong with my code, but I was concerned about what my boss would think when I sent him my finished product and it gave 11 compiler warnings.

Jeremy G · 04-18-2005, 01:52 PM

Quote:

Originally Posted by Salem

> gcc -o main.c main.exe
OMG!!!
So you want to OVERWRITE your source file with the compiled code?
Be more careful with your commands man, otherwise you'll be in the same position as another poor sap who managed to trash all his code a couple of hours before the "deadline".

It's a good job the command line was invalid as well, otherwise you'd be scrabbling round for a backup (you do have backups right?)

I use MSVC++. I copied this from some one who compiled my program with gcc for me to find out if there would be more/less warnings depending on compiler.

major_small · 04-18-2005, 07:50 PM

Quote:

Originally Posted by Jeremy G

I use MSVC++. I copied this from some one who compiled my program with gcc for me to find out if there would be more/less warnings depending on compiler.

he didn't use -Wall either

**lightatdawn** · 04-19-2005, 12:53 AM

WARNING: The following advice could be devastating in the hands of even a moderate coder. I must stress that 99.9% of warnings are valid and are possible sources of trouble. /WARNING

You could always use the warning specifier to remove specific undesirable warnings. I prefer to enable my compilers "warnings as errors" option, but should you be an experienced programmer you may safely be able to use the warning specifier.

#pragma warning ( disable, 4244 )

which would disable the "conversion from float to int, possible loss of data" warning.

Or even better: You can push and pop the warning state around the code that generates undesired warnings, and udjust the warning settings solely for that code or file. Check it out.

Peace

04-17-2005, 08:06 PM	#2
B0bDole Bob Dole for '08 Join Date: Sep 2004 Posts: 618	those warnings could cause memory problems... leaks, miss-assignments, etc. Something you probably dont see while doing tests on the program. __________________ Hmm
B0bDole is offline

04-17-2005, 08:33 PM	#3
gcn_zelda Rad Join Date: Mar 2003 Posts: 942	Yeah. Those kinds of warnings make programs like Windows come to pass. [/WindowsBash] __________________ -Stephen Cope smc42190@gmail.com http://purevolume.com/step http://myspace.com/stephencope
gcn_zelda is offline

04-17-2005, 08:52 PM	#4
Jeremy G mov.w #$1337,D0 Join Date: Nov 2001 Posts: 704	Both of you were probably right. Those warnings were actually pretty serious. And I've completely eliminated them. How ever, some warnings are just superflous. __________________ c++->visualc++->directx->opengl->c++; (it should be realized my posts are all in a light hearted manner. And should not be taken offense to.)
Jeremy G is offline

04-18-2005, 01:00 AM	#5
major_small Registered User Join Date: May 2003 Posts: 2,787	not really... the way I see it, you should try to minimize warnings by as much as possible... when I'm getting a warning, I find out what it means and if/how I can get rid of it. the standards were drawn up for a reason--there are reasons for every one of the warnings you get. it's not like they just said "let's throw this warning in there just for kicks". I agree--you can compile and run a program with no problems if you don't explicitly cast a float to an integer, but that doesn't help the readability of the code, and it may actually remind you that you're going to be losing data there... __________________ Join is in our Unofficial Cprog IRC channel Server: irc.phoenixradio.org Channel: #Tech Team Cprog Folding@Home: Team #43476 Download it Here Detailed Stats Here More Detailed Stats 52 Members so far, are YOU a member? Current team score: 1223226 (ranked 374 of 45152) The CBoard team is doing better than 99.16% of the other teams Top 5 Members: Xterria(518175), pianorain(118517), Bennet(64957), JaWiB(55610), alphaoide(44374) Last Updated on: Wed, 30 Aug, 2006 @ 2:30 PM EDT
major_small is offline

04-18-2005, 10:56 AM	#7
Salem and the hat of Jobseeking Join Date: Aug 2001 Location: The edge of the known universe Posts: 21,630	> gcc -o main.c main.exe OMG!!! So you want to OVERWRITE your source file with the compiled code? Be more careful with your commands man, otherwise you'll be in the same position as another poor sap who managed to trash all his code a couple of hours before the "deadline". It's a good job the command line was invalid as well, otherwise you'd be scrabbling round for a backup (you do have backups right?) __________________ If you dance barefoot on the broken glass of undefined behaviour, you've got to expect the occasional cut.
Salem is offline

04-18-2005, 12:50 PM	#8
skorman00 vae victus! Join Date: Nov 2003 Posts: 594	The superflous warnings you're talking about are probably like "unsigned being compared against signed value" or "float to int conversion, possible loss of data." For most pieces of data they won't matter, but they could come to bite you in the arse.
skorman00 is offline

04-18-2005, 01:01 PM	#9
joshdick Senior Member Join Date: Nov 2002 Location: Phildelphia, PA Posts: 1,146	The most superfluous warnings I've received were from Visual Studio 6.0 telling me that I had identifiers that were too long. I wasn't personally using 300-character identifiers. The problem is that a vector of strings expands out to an identifier of obscene length explictly stating the type of character set being used and the allocator that the vector is using and so on. Eleven warnings every compile like that, but not a single problem caused by it in testing, so meh. By the way, has anyone else ever run into that problem?
joshdick is offline

04-18-2005, 01:15 PM	#10
skorman00 vae victus! Join Date: Nov 2003 Posts: 594	I know a buddy who swears by long and descriptive function/variable names. He'd come up with something like Code: TakeTheXValueAndCalculateTheCorrectYValue(int thisIsTheXThatIsUsedToComputeTheY) <dramatization> I'm going to poke fun at him now with this, thanks joshdick. Some of you fellas/ladies may have already seen this, but it's always worth a second look. http://www.effect.net.au/lukastan/hu...ssages-Mac.htm
skorman00 is offline

04-19-2005, 12:53 AM	#15
lightatdawn Peace Join Date: Aug 2001 Posts: 1,512	WARNING: The following advice could be devastating in the hands of even a moderate coder. I must stress that 99.9% of warnings are valid and are possible sources of trouble. /WARNING You could always use the warning specifier to remove specific undesirable warnings. I prefer to enable my compilers "warnings as errors" option, but should you be an experienced programmer you may safely be able to use the warning specifier. #pragma warning ( disable, 4244 ) which would disable the "conversion from float to int, possible loss of data" warning. Or even better: You can push and pop the warning state around the code that generates undesired warnings, and udjust the warning settings solely for that code or file. Check it out. Peace __________________ "There's always another way" -lightatdawn (lightatdawn.cprogramming.com)
lightatdawn is offline

Thread Tools
Show Printable Version Email this Page
Display Modes
Linear Mode Switch to Hybrid Mode Switch to Threaded Mode

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
added start menu crashes game	avgprogamerjoe	Game Programming	6	08-29-2007 01:30 PM
Compiler Paths...	Cobra	C++ Programming	5	09-26-2006 04:04 AM
C Compiler and stuff	pal1ndr0me	C Programming	10	07-21-2006 11:07 AM
I can't get this new compiler to work.	Loduwijk	C++ Programming	7	03-29-2006 06:42 AM
how to call a compiler?	castlelight	C Programming	3	11-22-2005 11:28 AM