BugTraq
readdir_r considered harmful Nov 01 2005 03:57AM
Ben Hutchings (ben decadentplace org uk) (2 replies)
Re: readdir_r considered harmful Nov 05 2005 06:45PM
Casper Dik Sun COM (1 replies)
Re: [Full-disclosure] Re: readdir_r considered harmful Nov 06 2005 01:53AM
Ulrich Drepper (drepper gmail com) (1 replies)
On 11/5/05, Casper.Dik (at) sun (dot) com [email concealed] <Casper.Dik (at) sun (dot) com [email concealed]> wrote:
> Why not:
>
> 4. Require the readdir() implementation to use state local to dirp.
>
> I've never understood the rationale behind readdir_r;

Then you never really understood the implementation, seems. Of course
all implementations keep the content of the directory as read with
getdents or so in the DIR descriptor. But it is usually not the case
that the whole content fits into the buffer allocated. One could, of
course, resize the buffer to fit the content of the directory read,
even if this means reserving hundreds or thousands of kBs. But this
is not how most implementations work.

Instead implementations keep work similar to every buffered file I/O
operation. But this means that buffer content is replaced. If this
happens and some thread uses readdir() instead of readdir_r(), the
returned string pointer suddenly becomes invalid since it points to
memory which has been replaced.

Next time, before you make such comments, ask Don Cragun to explain
things to you.

[ reply ]
Re: [Full-disclosure] Re: readdir_r considered harmful Nov 06 2005 09:00AM
Casper Dik Sun COM (1 replies)
Re: [Full-disclosure] Re: readdir_r considered harmful Nov 06 2005 03:34PM
Ulrich Drepper (drepper gmail com) (1 replies)
Re: [Full-disclosure] Re: readdir_r considered harmful Nov 06 2005 07:19PM
Casper Dik Sun COM (1 replies)
Re: [Full-disclosure] Re: readdir_r considered harmful Nov 08 2005 01:47AM
Andrew Miller (andrew amxl com)
Re: readdir_r considered harmful Nov 01 2005 08:16PM
Ben Hutchings (ben decadentplace org uk)


 

Privacy Statement
Copyright 2010, SecurityFocus