Bugs item #1579171, was opened at 2006-10-17 20:29
Message generated for change (Comment added) made by bagder
You can respond by visiting:
https://sourceforge.net/tracker/?func=detail&atid=100976&aid=1579171&group_id=976

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: debug/info output
Group: None
>Status: Closed
>Resolution: Fixed
Priority: 5
Submitted By: tyrolean (frzem)
Assigned to: Daniel Stenberg (bagder)
Summary: Code-analyze (with prefast)

Initial Comment:
I've build analyze-report with prefast (from
Microsoft's DDK), and this tool found following
(potential) problems:

* against released source-code 7.15.5

warning 8132 : Taking the size of pointer tsd: This
will yield the size of a pointer (4 or 8), not the size
of the object pointed to. Dereference the pointer, or
if the size of a pointer was intended, use the pointer
type or (void *) instead.
             File path: curl-7.15.5\lib\hostthre.c
             Function: init_thread_sync_data
             Line: 211
memset(tsd, 0, sizeof(tsd));

warning 11 : Dereferencing NULL pointer 'proxy_password'.
             File path: curl-7.15.5\lib\url.c
             Function: handleSock5Proxy
             Line: 2151
memcpy(socksreq + len, proxy_password, (int) pwlen);

Most of the other warnings are on isdigit(), isspace()
..., because MS (Visual Studio) declared the parameter-
type as unsigned char, and the macro-implementation of
these functions have a wrong behaviour with ASCII-
Codes above 0x80 (because they use a array of flags).

BG,
FRZEM

----------------------------------------------------------------------

>Comment By: Daniel Stenberg (bagder)
Date: 2006-10-17 23:34

Message:
Logged In: YES
user_id=1110

Fixed and committed the sdigit()/isspace() range of problems
as well.

----------------------------------------------------------------------

Comment By: Daniel Stenberg (bagder)
Date: 2006-10-17 22:53

Message:
Logged In: YES
user_id=1110

Thanks!

Please do a fresh test against the CVS version (or a recent
snapshot) as lots of things have changed since 7.15.5.

Your first flaw is now fixed, the socks5 one is a false
positive and

I'm working on the is*() case with negative number input, as
it seems to be a correct report that indeed the code suffers
from all over.

----------------------------------------------------------------------

Comment By: tyrolean (frzem)
Date: 2006-10-17 20:45

Message:
Logged In: YES
user_id=1623281

Sorry, but with the isdigit(), isspace() ... - problem(s),
I mean, that the macros have problems with a negative
int-value !

BG,
FRZEM

----------------------------------------------------------------------

You can respond by visiting:
https://sourceforge.net/tracker/?func=detail&atid=100976&aid=1579171&group_id=976
Received on 2006-10-17