Re: [PATCH] One more -x to not traverse cyclic bind-mounts

2011-01-19 Thread Jeff Hansen
On 01/18/2011 05:45 PM, Wayne Davison wrote: On Tue, Jan 18, 2011 at 11:27 AM, Jeff Hansen jhan...@cardaccess-inc.com mailto:jhan...@cardaccess-inc.com wrote: I hate to beat an (apparently) dead horse, but I've tested this patch on several systems and I would really appreciate it if

Re: [PATCH] One more -x to not traverse cyclic bind-mounts

2011-01-19 Thread Matt McCutchen
On Wed, 2011-01-19 at 12:40 -0700, Jeff Hansen wrote: The way rsync determines if the directory is a mount point looks very portable to me (with the FLAG_TOP_DIR bit), and with this patch, it has worked on every Linux system I've tried so far (not sure if FLAG_TOP_DIR works on BSDs, but I'm

Re: [PATCH] One more -x to not traverse cyclic bind-mounts

2011-01-19 Thread Jeff Hansen
On 01/19/2011 12:52 PM, Matt McCutchen wrote: On Wed, 2011-01-19 at 12:40 -0700, Jeff Hansen wrote: The way rsync determines if the directory is a mount point looks very portable to me (with the FLAG_TOP_DIR bit), and with this patch, it has worked on every Linux system I've tried so far (not

Re: [PATCH] One more -x to not traverse cyclic bind-mounts

2011-01-18 Thread Jeff Hansen
On 01/13/2011 06:14 PM, Wayne Davison wrote: On Thu, Jan 13, 2011 at 10:24 AM, Jeff Hansen jhan...@cardaccess-inc.com mailto:jhan...@cardaccess-inc.com wrote: This patch allows one to repeat the '-x' option a third time to stop rsync from traversing potentially cyclic bind mounts that

Re: [PATCH] One more -x to not traverse cyclic bind-mounts

2011-01-18 Thread Wayne Davison
On Tue, Jan 18, 2011 at 11:27 AM, Jeff Hansen jhan...@cardaccess-inc.comwrote: I hate to beat an (apparently) dead horse, but I've tested this patch on several systems and I would really appreciate it if this patch (or some variation of it) went into mainline. You patch just prevents

Re: [PATCH] One more -x to not traverse cyclic bind-mounts

2011-01-14 Thread Jeff Hansen
On 01/13/2011 06:14 PM, Wayne Davison wrote: On Thu, Jan 13, 2011 at 10:24 AM, Jeff Hansen jhan...@cardaccess-inc.com mailto:jhan...@cardaccess-inc.com wrote: This patch allows one to repeat the '-x' option a third time to stop rsync from traversing potentially cyclic bind mounts that

[PATCH] One more -x to not traverse cyclic bind-mounts

2011-01-13 Thread Jeff Hansen
This patch allows one to repeat the '-x' option a third time to stop rsync from traversing potentially cyclic bind mounts that are on the same filesystem. For example, the following scenario would cyclically traverse the bind mounts until the OS (Linux in this case) hits a maximum traversal

Re: [PATCH] One more -x to not traverse cyclic bind-mounts

2011-01-13 Thread Matt McCutchen
On Thu, 2011-01-13 at 11:24 -0700, Jeff Hansen wrote: This patch allows one to repeat the '-x' option a third time to stop rsync from traversing potentially cyclic bind mounts that are on the same filesystem. Your patch does not do anything about bind mounts, it simply excludes subdirectories

Re: [PATCH] One more -x to not traverse cyclic bind-mounts

2011-01-13 Thread Wayne Davison
On Thu, Jan 13, 2011 at 10:24 AM, Jeff Hansen jhan...@cardaccess-inc.comwrote: This patch allows one to repeat the '-x' option a third time to stop rsync from traversing potentially cyclic bind mounts that are on the same filesystem. See the support/mnt-excl script for one way exclude all