Re: [PATCH v2 07/11] CryptoPkg: Clean-up CRT Library Wrapper.

["Long, Qin" <qin.long@intel.com>]

Laszlo,

Thanks for catching this.

I was using GCC5 toolchain for few validations, the build looks OK. 
Yes, should have more toolchain validations. :-(

I am also confused about this nonull warning.  Maybe strncasecmp function 
could have been defined as the different prototype (explicitely or implicitely):
	int strcasecmp(const char *s1, const char *s2) __attribute__((nonnull));

And there was also one extra build warning (on openssl source) under GCC48.

Let me dig more and provide possible fixes to address this later.


Best Regards & Thanks,
LONG, Qin


> -----Original Message-----
> From: edk2-devel [mailto:edk2-devel-bounces@lists.01.org] On Behalf Of
> Laszlo Ersek
> Sent: Friday, March 31, 2017 1:34 AM
> To: Long, Qin; edk2-devel@lists.01.org; Gerd Hoffmann
> Cc: Ye, Ting; ronald.cron@arm.com; Wu, Jiaxin; glin@suse.com;
> ard.biesheuvel@linaro.org
> Subject: Re: [edk2] [PATCH v2 07/11] CryptoPkg: Clean-up CRT Library
> Wrapper.
> 
> Qin,
> 
> this patch (commit fc9fa685d689c) seems to break *some* GCC builds:
> 
> On 03/23/17 14:19, Qin Long wrote:
> 
> [snip]
> 
> > +/* Compare first n bytes of string s1 with string s2, ignoring case
> > +*/ int strncasecmp (const char *s1, const char *s2, size_t n) {
> > +  int Val;
> > +
> > +  ASSERT(s1 != NULL);
> > +  ASSERT(s2 != NULL);
> 
> Here. The error messages are:
> 
>   CryptoPkg/Library/BaseCryptLib/SysCall/CrtWrapper.c: In function
> 'strncasecmp':
>   MdePkg/Include/Library/DebugLib.h:292:12: error: nonnull argument 's1'
> compared to NULL [-Werror=nonnull-compare]
>            if (!(Expression)) {        \
>             ^
>   CryptoPkg/Library/BaseCryptLib/SysCall/CrtWrapper.c:144:3: note: in
> expansion of macro 'ASSERT'
>      ASSERT(s1 != NULL);
>      ^~~~~~
>   MdePkg/Include/Library/DebugLib.h:292:12: error: nonnull argument 's2'
> compared to NULL [-Werror=nonnull-compare]
>            if (!(Expression)) {        \
>               ^
>   CryptoPkg/Library/BaseCryptLib/SysCall/CrtWrapper.c:145:3: note: in
> expansion of macro 'ASSERT'
>      ASSERT(s2 != NULL);
>      ^~~~~~
> 
> The GCC documentation <https://gcc.gnu.org/onlinedocs/gcc/Warning-
> Options.html> says,
> 
>   -Wnonnull-compare
> 
>       Warn when comparing an argument marked with the nonnull function
> attribute against null inside the function.
> 
>       -Wnonnull-compare is included in -Wall. It can be disabled with the -Wno-
> nonnull-compare option.
> 
> I have no idea where the compiler thinks "s1" and "s2" are marked as non-
> null. I grepped the tree for "nonnull" and there are no  hits.
> 
> Gerd, what compiler are you using? (From your logs, it looks like OVMF is
> being built for IA32, with the GCC49 tool chain settings.)
> 
> Thanks
> Laszlo
> 
> _______________________________________________
> edk2-devel mailing list
> edk2-devel@lists.01.org
> https://lists.01.org/mailman/listinfo/edk2-devel