[syslinux] [PATCH 1/1] : PCI detection code doesn't corrupt memory anymore

[H. Peter Anvin]

Erwan wrote:
>  From : Erwan Velu <erwan.velu at free.fr>
> 
> This patch fixes the pci detection stuff demo'd in pcitest.c32
> - Fix a wrong type in malloc
> - Incorrect use of strdup
> 
> This two fixes makes pcitest.c32 working far better, no more crash when 
> running it twice.
> 
> I'd like to thank Sebastian Herbszt for its reports and tests.
> 
> This patch was made for 3.62-pre16 but applies fine on 3.63-pre1.
> 

I notice you use strncpy() in a number of places here.  strncpy() works 
differently from just about all the other strn*() functions, in that it 
doesn't make sure that the final output is zero-terminated.  strncpy() 
really is mostly to ensure that binary output doesn't leak information, 
i.e. it produces *zero-padded* output (as opposed to zero-terminated.)

If you're looking for something that's equivalent to
"snprintf(dst, dst_len, "%s", src)" that's strlcpy() [which I'm not sure 
if I have in libcom32, but can be trivially added if needed.]

	-hpa