Re: using access(2)

2012-03-23 Thread Ted Unangst
On Fri, Mar 23, 2012, Jan Stary wrote: > However, access() is used in test(1): > > Is this a problem? It is not possible to close the race between when test sees a file and when the parent process sees test's exit code. Write your shell scripts with caution.

Re: using access(2)

2012-03-23 Thread Pieter Verberne
On 2012-03-23 21:04, Theo de Raadt wrote: The manpage of access(2) says CAVEATS access() and faccessat() should never be used for actual access control. Doing so can result in a time of check vs. time of use security hole. However, access() is used in test(1):

Re: using access(2)

2012-03-23 Thread Theo de Raadt
> The manpage of access(2) says > > CAVEATS > access() and faccessat() should never be used for actual access > control. Doing so can result in a time of check vs. time of use > security hole. > > However, access() is used in test(1): > > case FILRD: > retu

using access(2)

2012-03-23 Thread Jan Stary
The manpage of access(2) says CAVEATS access() and faccessat() should never be used for actual access control. Doing so can result in a time of check vs. time of use security hole. However, access() is used in test(1): case FILRD: return access(nm