On Thu, Feb 24, 2000 at 11:31:43AM -0500, Alan Curry wrote:
> Harald Kirsch writes the following:
> >The point is that on all other *nixes (as reported in the Tcl-newsgroup)
> >the behaviour is different. getcwd does return `permission denied' if
> >the x-flag of the current dir or any of its parents is not present.
> >
> >But maybe all other systems are not POSIX and Linux is or Linux knows
> >better than POSIX?
>
> You might get a better response if you can point to the POSIX requirement
> that getcwd() behave specifically as you describe, or that the system is not
> allowed to have additional syscalls that provide slightly more information
> than is available through opendir/readdir/etc.
Ok, a fair request and I can not currently cite the part of POSIX requiring
the behaviour I describe.
On the other hand, as Jamie.Lokier@cern.ch explained:
> To be more precise, read permission controls whether you can readdir(),
> and execute permission controls whether you can lookup entries e.g. to
> stat(), open(), readlink() or follow subdirectory etc.
Your reasoning would mean that we can implement `additional syscalls'
linuxstat(), linuxopen() and linuxreadlink() which `provide slightly
more information', e.g. by not at all looking at permissions.
Of course we can, but should we?
Harald Kirsch
-- -------------------------------------------------+------------------ Harald Kirsch, kir@iitb.fhg.de, +49 721 6091 369 | FhG/IITB, Fraunhoferstr.1, 76131 Karlsruhe |- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Tue Feb 29 2000 - 21:00:10 EST