linux-mips
[Top] [All Lists]

Re: Patches for 34K APRP

To: Ralf Baechle <ralf@linux-mips.org>
Subject: Re: Patches for 34K APRP
From: "Kevin D. Kissell" <kevink@mips.com>
Date: Thu, 17 Apr 2008 15:43:02 +0200
Cc: linux-mips@linux-mips.org
In-reply-to: <20080417132610.GB31453@linux-mips.org>
Original-recipient: rfc822;linux-mips@linux-mips.org
References: <4805FFE6.5080903@mips.com> <20080417132610.GB31453@linux-mips.org>
Sender: linux-mips-bounce@linux-mips.org
User-agent: Thunderbird 2.0.0.12 (Windows/20080213)
Ralf Baechle wrote:
On Wed, Apr 16, 2008 at 03:32:22PM +0200, Kevin D. Kissell wrote:

--- a/arch/mips/kernel/kspd.c
+++ b/arch/mips/kernel/kspd.c
@@ -257,7 +257,7 @@ void sp_work_handle_request(void)
vcwd = vpe_getcwd(tclimit); - /* change to the cwd of the process that loaded the SP program */
+               /* change to cwd of the process that loaded the SP program */
                old_fs = get_fs();
                set_fs(KERNEL_DS);
                sys_chdir(vcwd);
@@ -323,6 +323,9 @@ static void sp_cleanup(void)
                        set >>= 1;
                }
        }
+
+       /* Put daemon cwd back to root to avoid umount problems */
+       sys_chdir("/");

Still not kosher; there might be multiple roots on a system ...
What do you suggest, then?  Leaving things as they are makes it impossible
to do a clean shutdown of an APRP system if the kspd was last used to support an RP binary that was read from a mounted file system. I respectfully submit that it's at least an order of magnitude more likely that a mounted file system
will have been used than a chrooted, mounted file system.

         Regards,

         Kevin K.

<Prev in Thread] Current Thread [Next in Thread>