weisen@eniac.seas.upenn.edu) (Neil I. Weisenfeld ;-) (12/23/89)
It's been a while since this happened so I don't have the code lying
around, but....
I am using gcc v1.35 under DEC's ULTRIX. I was diligently working on
a project that was due the next morning when quite a bit of my time
was taken up trying to find a really weird bug. As it turns out, I
misspelled `default' as `defualt' in a switch statement. The code
compiled without an error, however the `defualt' was never executed;
it just dropped out of the switch. Is this a bug or is there a
rational `C' explanation for this?
Neil
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
# ## Neil Weisenfeld | Internet:
### # Univ. of Pennsylvania | weisen@eniac.seas.upenn.edu
# # Computer Science Dept. |
##### # Class of 1991 |
# # | USPS:
### # <-giant killer | 3700 Spruce St., Box 572
# ## net-smile | Philadelphia, PA 19104
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-cpcahil@virtech.uucp (Conor P. Cahill) (12/23/89)
In article <18472@netnews.upenn.edu>, weisen@eniac.seas.upenn.edu) (Neil I. Weisenfeld ;-) writes: > I am using gcc v1.35 under DEC's ULTRIX. I was diligently working on > a project that was due the next morning when quite a bit of my time > was taken up trying to find a really weird bug. As it turns out, I > misspelled `default' as `defualt' in a switch statement. The code > compiled without an error, however the `defualt' was never executed; > it just dropped out of the switch. Is this a bug or is there a > rational `C' explanation for this? This will happen in any C compiler. Since the defualt does not match "default" it is then interpreted as a label (used by goto's). So your defualt: code would be executed as part of the preceding case. If you had linted the code, you would have gotten a message about defualt being unused. This would have pointed out the problem to you. -- +-----------------------------------------------------------------------+ | Conor P. Cahill uunet!virtech!cpcahil 703-430-9247 ! | Virtual Technologies Inc., P. O. Box 876, Sterling, VA 22170 | +-----------------------------------------------------------------------+
weisen@eniac.seas.upenn.edu) (Neil I. Weisenfeld ;-) (12/24/89)
Thanks to all those who gave me (non-obnoxious) replies to my posting.
I didn't realize that gcc would interpret `defualt' in a case xxx: as
a label if it hadn't been defined. Is this non-ANSI C not to give an
error for undefined and un-prototyped symbols.
Neil
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
# ## Neil Weisenfeld | Internet:
### # Univ. of Pennsylvania | weisen@eniac.seas.upenn.edu
# # Computer Science Dept. |
##### # Class of 1991 |
# # | USPS:
### # <-giant killer | 3700 Spruce St., Box 572
# ## net-smile | Philadelphia, PA 19104
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-cpcahil@virtech.uucp (Conor P. Cahill) (12/24/89)
In article <18485@netnews.upenn.edu>, weisen@eniac.seas.upenn.edu) (Neil I. Weisenfeld ;-) writes: > Thanks to all those who gave me (non-obnoxious) replies to my posting. > I didn't realize that gcc would interpret `defualt' in a case xxx: as > a label if it hadn't been defined. Is this non-ANSI C not to give an > error for undefined and un-prototyped symbols. How does one define or prototype a statement label? Interpreting "defualt" as a statement label is correct in all versions of C. -- +-----------------------------------------------------------------------+ | Conor P. Cahill uunet!virtech!cpcahil 703-430-9247 ! | Virtual Technologies Inc., P. O. Box 876, Sterling, VA 22170 | +-----------------------------------------------------------------------+
tiemann@hudson.cs.columbia.edu (Michael Tiemann) (12/27/89)
The ratinal explanatin os that the compiler interpreted `defualt:' as a valid label in the function. -Wswitch might have helped you if you were using enumerated values in the switch. Michael