[Cluster-devel] [PATCH] dlm_controld: fix plock dev_write no op

[Cluster-devel] [PATCH] dlm_controld: fix plock dev_write no op

[David Teigland]

When a plock unlock is received due to the file
being closed (the CLOSE flag is set), we should
not write an unlock result back to the kernel.
If we do, the kernel, which does not expect a
reply, will report the error "dev_write no op".

In cases where dlm_controld encounters and error
handling the unlock operation, it was writing
the error result back to the kernel, even though
the unlock was flagged with CLOSE.  The fix is
to check for the CLOSE flag and skip writing
the error result, as we do with normal results.

This problem is especially visible when using
flocks (not plocks).  This is because the kernel
generates extraneous plock unlock requests
when files are closed with flocks.  Because
dlm_controld finds no plocks on the files,
it replies to the kernel with an error, rather
than skipping the reply to do CLOSE.

bz 731775

Signed-off-by: David Teigland <teigland@redhat.com>
---
 group/dlm_controld/plock.c |    6 ++++--
 1 files changed, 4 insertions(+), 2 deletions(-)

diff --git a/group/dlm_controld/plock.c b/group/dlm_controld/plock.c
index 6d5dea8..556993b 100644
--- a/group/dlm_controld/plock.c
+++ b/group/dlm_controld/plock.c
@@ -1583,8 +1583,10 @@ void process_plocks(int ci)
 	return;
 
  fail:
-	info.rv = rv;
-	rv = write(plock_device_fd, &info, sizeof(info));
+	if (!(info.flags & DLM_PLOCK_FL_CLOSE)) {
+		info.rv = rv;
+		rv = write(plock_device_fd, &info, sizeof(info));
+	}
 }
 
 void process_saved_plocks(struct lockspace *ls)
-- 
1.7.6

[Cluster-devel] [PATCH] dlm_controld: fix plock dev_write no op

[Fabio M. Di Nitto]

The patch looks simple and safe. Does it apply only to rhel6 or also
upstream?

Fabio

On 8/18/2011 9:38 PM, David Teigland wrote:
> When a plock unlock is received due to the file
> being closed (the CLOSE flag is set), we should
> not write an unlock result back to the kernel.
> If we do, the kernel, which does not expect a
> reply, will report the error "dev_write no op".
> 
> In cases where dlm_controld encounters and error
> handling the unlock operation, it was writing
> the error result back to the kernel, even though
> the unlock was flagged with CLOSE.  The fix is
> to check for the CLOSE flag and skip writing
> the error result, as we do with normal results.
> 
> This problem is especially visible when using
> flocks (not plocks).  This is because the kernel
> generates extraneous plock unlock requests
> when files are closed with flocks.  Because
> dlm_controld finds no plocks on the files,
> it replies to the kernel with an error, rather
> than skipping the reply to do CLOSE.
> 
> bz 731775
> 
> Signed-off-by: David Teigland <teigland@redhat.com>
> ---
>  group/dlm_controld/plock.c |    6 ++++--
>  1 files changed, 4 insertions(+), 2 deletions(-)
> 
> diff --git a/group/dlm_controld/plock.c b/group/dlm_controld/plock.c
> index 6d5dea8..556993b 100644
> --- a/group/dlm_controld/plock.c
> +++ b/group/dlm_controld/plock.c
> @@ -1583,8 +1583,10 @@ void process_plocks(int ci)
>  	return;
>  
>   fail:
> -	info.rv = rv;
> -	rv = write(plock_device_fd, &info, sizeof(info));
> +	if (!(info.flags & DLM_PLOCK_FL_CLOSE)) {
> +		info.rv = rv;
> +		rv = write(plock_device_fd, &info, sizeof(info));
> +	}
>  }
>  
>  void process_saved_plocks(struct lockspace *ls)