Re: New patch for bdflush and syslog

From: jmjonesat_private
Date: Sat Sep 01 2001 - 15:12:15 PDT

  • Next message: jmjonesat_private: "Re: quotactl hook"

    On Fri, 31 Aug 2001, Wayne Salamon wrote:
    
    >   In the case of this hook, 'data' is either a value (5, 10, 35, etc.), or
    > a pointer to user-space memory where values are written. If somebody is
    > writing a security module, they better know what the parameters mean. The
    > question is, would an LSM module ever have to write into the user data
    > space?  Well, based on the placement of the hook call, no, because the
    > kernel will write the data after the hook is called.
    > 
    >   I don't see a race condition either, because the data read is by value.
    > What types of dangererous conditions do you think this hook creates?
    
    I can't identify a convincing race condition in this case, either,
    although I admit I'm not a "genius" in that regard... still thinking in
    terms of one-processor/one-task-at-a-time.   Could somebody point
    out a specific example of where it might be a problem?
    
    > Wayne Salamon 
    > wsalamonat_private
    
    J. Melvin Jones
    
    |>------------------------------------------------------
    ||  J. MELVIN JONES            jmjonesat_private 
    |>------------------------------------------------------
    ||  Microcomputer Systems Consultant  
    ||  Software Developer
    ||  Web Site Design, Hosting, and Administration
    ||  Network and Systems Administration
    |>------------------------------------------------------
    ||  http://www.jmjones.com/
    |>------------------------------------------------------
    
    
    
    
    _______________________________________________
    linux-security-module mailing list
    linux-security-moduleat_private
    http://mail.wirex.com/mailman/listinfo/linux-security-module
    



    This archive was generated by hypermail 2b30 : Sat Sep 01 2001 - 15:13:43 PDT