Using GCC 2.95.3 on FreeBSD 4.7 (which has a couple of tests not usually supported by GCC 2.95) I get a couple of warnings due to recent changes/additions in Wine; these warnings seem to indicate real problems:
rtlstr.c: In function `test_RtlInitUnicodeString': rtlstr.c:110: warning: implicit declaration of function `wcscmp' rtlstr.c: In function `test_RtlUnicodeStringToInteger': rtlstr.c:147: warning: decimal constant is so large that it is unsigned rtlstr.c:148: warning: decimal constant is so large that it is unsigned rtlstr.c:170: warning: int format, different type arg (arg 3) rtlstr.c:175: warning: int format, different type arg (arg 3) rtlstr.c:176: warning: int format, different type arg (arg 3) rtlstr.c:178: warning: int format, different type arg (arg 3) rtlstr.c:179: warning: int format, different type arg (arg 3) rtlstr.c:181: warning: int format, different type arg (arg 3) rtlstr.c:182: warning: int format, different type arg (arg 3) rtlstr.c:184: warning: int format, different type arg (arg 3) rtlstr.c:185: warning: int format, different type arg (arg 3)
Robert, perhaps you could have a look at these?
Gerald
Gerald Pfeifer wrote:
Using GCC 2.95.3 on FreeBSD 4.7 (which has a couple of tests not usually supported by GCC 2.95) I get a couple of warnings due to recent changes/additions in Wine; these warnings seem to indicate real problems:
I brought those up recently; Francois Gouget suggests fixing the wcscmp warning by switching to strcmpW http://www.winehq.com/hypermail/wine-devel/2002/12/0892.html
I fixed the decimal constant warnings in a patch I sent a couple days ago... but my mail system seems to have dropped it, so I'll resend it.
The format warnings are worrisome.
rtlstr.c: In function `test_RtlInitUnicodeString': rtlstr.c:110: warning: implicit declaration of function `wcscmp' rtlstr.c: In function `test_RtlUnicodeStringToInteger': rtlstr.c:147: warning: decimal constant is so large that it is unsigned rtlstr.c:148: warning: decimal constant is so large that it is unsigned rtlstr.c:170: warning: int format, different type arg (arg 3) rtlstr.c:175: warning: int format, different type arg (arg 3) rtlstr.c:176: warning: int format, different type arg (arg 3) rtlstr.c:178: warning: int format, different type arg (arg 3) rtlstr.c:179: warning: int format, different type arg (arg 3) rtlstr.c:181: warning: int format, different type arg (arg 3) rtlstr.c:182: warning: int format, different type arg (arg 3) rtlstr.c:184: warning: int format, different type arg (arg 3) rtlstr.c:185: warning: int format, different type arg (arg 3)
Robert, perhaps you could have a look at these?
Gerald
Sent a fix for wcscmp vs. lstrcmpW.
On Fri, 27 Dec 2002, Dan Kegel wrote: [...]
rtlstr.c:170: warning: int format, different type arg (arg 3) rtlstr.c:175: warning: int format, different type arg (arg 3) rtlstr.c:176: warning: int format, different type arg (arg 3) rtlstr.c:178: warning: int format, different type arg (arg 3) rtlstr.c:179: warning: int format, different type arg (arg 3) rtlstr.c:181: warning: int format, different type arg (arg 3) rtlstr.c:182: warning: int format, different type arg (arg 3) rtlstr.c:184: warning: int format, different type arg (arg 3) rtlstr.c:185: warning: int format, different type arg (arg 3)
'ok' uses the standard glibc printf functions and thus does not support '%S' (unicode strings). Or if it does support '%S' it expects 32bit Unicode chars rather than 16bit ones so it won't work.
Unfortunately, currently there is no good way of printing Unicode strings in tests. We may need a function similar to debugstr_w...
-
Dan Kegel -
Francois Gouget -
Gerald Pfeifer