qemu-ppc
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Qemu-ppc] [PATCH] fdt_ro.c: implement strnlen


From: David Gibson
Subject: Re: [Qemu-ppc] [PATCH] fdt_ro.c: implement strnlen
Date: Thu, 19 Oct 2017 15:11:34 +1100
User-agent: Mutt/1.9.1 (2017-09-22)

On Wed, Oct 18, 2017 at 10:39:57PM -0400, Programmingkid wrote:
> 
> > On Oct 18, 2017, at 9:31 PM, David Gibson <address@hidden> wrote:
> > 
> > On Wed, Oct 18, 2017 at 06:31:16PM -0400, John Arbuckle wrote:
> >> Implement the strnlen() function if it isn't implemented.
> >> 
> >> Signed-off-by: John Arbuckle <address@hidden>
> > 
> > Nice idea, but this won't work.
> > 
> >> ---
> >> libfdt/fdt_ro.c | 24 ++++++++++++++++++++++++
> >> 1 file changed, 24 insertions(+)
> >> 
> >> diff --git a/libfdt/fdt_ro.c b/libfdt/fdt_ro.c
> >> index 3d00d2e..a7986fb 100644
> >> --- a/libfdt/fdt_ro.c
> >> +++ b/libfdt/fdt_ro.c
> >> @@ -55,6 +55,30 @@
> >> 
> >> #include "libfdt_internal.h"
> >> 
> >> +/* if the current environment does not define strnlen */
> >> +#ifndef strnlen
> > 
> > This will only trigger if strnlen is defined as a macro.  The C
> > library (or other environment) *might* do that, but there's no
> > guarantee, whether or not it defines strnlen as a function.  With
> > extra complications if the compiler has a strnlen builtin like gcc.
> 
> :(
> 
> >> +
> >> +/* This eliminates the missing prototype warning */
> >> +int strnlen(const char *string, int max_count);
> >> +
> >> +/*
> >> + * strnlen: return the length of a string or max_count
> >> + * which ever is shortest
> >> + */
> >> +
> >> +int strnlen(const char *string, int max_count)
> > 
> > Also strnlen is supposed to take and return size_t, not int.
> 
> Will change this.
> 
> > 
> >> +{
> >> +    int count;
> >> +    for(count = 0; count < max_count; count++) {
> >> +        if (string[count] == '\0') {
> >> +            break;
> >> +        }
> >> +    }
> >> +    return count;
> >> +}
> >> +
> >> +#endif /* strnlen */
> >> +
> >> static int _fdt_nodename_eq(const void *fdt, int offset,
> >>                        const char *s, int len)
> >> {
> > 
> > In any case, this sort of compatibility munging is the job of
> > libfdt_env.h.  It's purpose is to provide all the external things that
> > libfdt needs - there's not that many of them and strnlen() is one.
> > The included libfdt_env.h is intended for POSIXy userspace
> > environments, which should already include strnlen() in string.h.
> 
> QEMU supports Mac OS 10.5 and higher. The strnlen() function might
> have been added to Mac OS X in version 10.7.

Wow.  No strnlen() in MacOS until that recently, that's pretty shit.

> libfdt prevents QEMU
> from compiling on Mac OS X because of the missing strnlen()
> function. This issue needs to be resolved. If libfdt_env.h is where
> you want a fix located, would something like this be good:

> 
> #ifdef __APPLE__
> 
> #if (MAC_OS_X_VERSION_MAX_ALLOWED < MAC_OS_X_VERSION_10_7)
>       size_t strnlen(const char *string, int max_count)
>       {
>               ... // pretend this is implemented
>       }
> #endif 
> 
> #endif

I'm not sure I'd go as far as to call it "good" - it's kinda ugly -
but it would be an acceptable solution (as long as max_count is
changed to size_t as well, of course).

-- 
David Gibson                    | I'll have my music baroque, and my code
david AT gibson.dropbear.id.au  | minimalist, thank you.  NOT _the_ _other_
                                | _way_ _around_!
http://www.ozlabs.org/~dgibson

Attachment: signature.asc
Description: PGP signature


reply via email to

[Prev in Thread] Current Thread [Next in Thread]