[PATCH] scsi: fix async probe regression

[PATCH] scsi: fix async probe regression

[Dan Williams]

Commit a7a20d1 "[SCSI] sd: limit the scope of the async probe domain"
moved sd probe work out of reach of wait_for_device_probe().  Allow it
to be synced via scsi_complete_async_scans().

Cc: David Miller <davem@davemloft.net>
Cc: Alan Stern <stern@rowland.harvard.edu>
Reported-by: Meelis Roos <mroos@linux.ee>
Tested-by: Meelis Roos <mroos@linux.ee>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
---

 So this is the stop gap that allows us to keep a7a20d1 as a fix for a
 resume deadlock, and gives us time to figure out how to clarify
 async_synchronize_full() properly in the 3.6 timeframe.

 drivers/scsi/scsi_scan.c |   12 ++++++++++--
 1 file changed, 10 insertions(+), 2 deletions(-)

diff --git a/drivers/scsi/scsi_scan.c b/drivers/scsi/scsi_scan.c
index 01b0374..41f9192 100644
--- a/drivers/scsi/scsi_scan.c
+++ b/drivers/scsi/scsi_scan.c
@@ -141,13 +141,13 @@ struct async_scan_data {
  * started scanning after this function was called may or may not have
  * finished.
  */
-int scsi_complete_async_scans(void)
+static void __scsi_complete_async_scans(void)
 {
 	struct async_scan_data *data;
 
 	do {
 		if (list_empty(&scanning_hosts))
-			return 0;
+			return;
 		/* If we can't get memory immediately, that's OK.  Just
 		 * sleep a little.  Even if we never get memory, the async
 		 * scans will finish eventually.
@@ -181,6 +181,14 @@ int scsi_complete_async_scans(void)
 	spin_unlock(&async_scan_lock);
 
 	kfree(data);
+}
+
+
+int scsi_complete_async_scans(void)
+{
+	__scsi_complete_async_scans();
+	async_synchronize_full_domain(&scsi_sd_probe_domain);
+
 	return 0;
 }
 

Re: [PATCH] scsi: fix async probe regression

[James Bottomley]

On Fri, 2012-05-25 at 09:39 -0700, Dan Williams wrote:
> Commit a7a20d1 "[SCSI] sd: limit the scope of the async probe domain"
> moved sd probe work out of reach of wait_for_device_probe().  Allow it
> to be synced via scsi_complete_async_scans().
> 
> Cc: David Miller <davem@davemloft.net>
> Cc: Alan Stern <stern@rowland.harvard.edu>
> Reported-by: Meelis Roos <mroos@linux.ee>
> Tested-by: Meelis Roos <mroos@linux.ee>
> Signed-off-by: Dan Williams <dan.j.williams@intel.com>
> ---
> 
>  So this is the stop gap that allows us to keep a7a20d1 as a fix for a
>  resume deadlock, and gives us time to figure out how to clarify
>  async_synchronize_full() properly in the 3.6 timeframe.
> 
>  drivers/scsi/scsi_scan.c |   12 ++++++++++--
>  1 file changed, 10 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/scsi/scsi_scan.c b/drivers/scsi/scsi_scan.c
> index 01b0374..41f9192 100644
> --- a/drivers/scsi/scsi_scan.c
> +++ b/drivers/scsi/scsi_scan.c
> @@ -141,13 +141,13 @@ struct async_scan_data {
>   * started scanning after this function was called may or may not have
>   * finished.
>   */
> -int scsi_complete_async_scans(void)
> +static void __scsi_complete_async_scans(void)

Rather than wrappering the function, can we just fix it?

Put the async_synchronize_full_domain(&scsi_sd_probe_domain); at the
bottom with an out: label and make the only other return in the function
become a goto out;

Thanks,

James