: SCSI: fix /proc memory leak in the SCSI core
Tony Battersby
Tested-byI tested Alan's original patch with sym53c8xx back in January:
http://marc.info/?l=linux-scsi&m=123143710010939&w=4
Tested-by Tony Battersby <[email protected]>
Quoted 5 lines · Tony Battersby
I tested Alan's original patch with sym53c8xx back in January:
http://marc.info/?l=linux-scsi&m=123143710010939&w=4
Tested-by Tony Battersby <[email protected]>
Thanks for the support. I must have missed your post when it appeared last month.
In fact, I don't care which approach gets accepted: James's or my own. His has the advantage of being somewhat shorter and not changing the order of events during host creation.
On the other hand, James, your suggested patch is definitely wrong, for a trivial reason. You moved the scsi_proc_hostdir_rm() call into scsi_host_put() instead of scsi_host_dev_release(), where it belongs. I'll post a corrected version shortly.
Alan Stern