On Wed, May 29, 2019 at 02:13:18PM -0700, David Rientjes wrote: > On Wed, 29 May 2019, Hariprasad Kelam wrote: > > > dont acquire lock before calling wd719x_chip_init. > > > > Issue identified by coccicheck > > > > Signed-off-by: Hariprasad Kelam <hariprasad.kelam@xxxxxxxxx> > > ----- > > changes in v1: Replace GFP_KERNEL with GFP_ATOMIC. > > changes in v2: Call wd719x_chip_init without lock as suggested > > in review > > Why was host_lock taken here initially? I assume it's to protect some > race in init that leads to an undefined state. wd719x_chip_init is getting called from wd719x_host_reset and wd719x_board_found. In wd719x_board_found case its not acquiring any lock. In wd719x_host_reset it is called under spin_lock. Acquiring spin_lock in wd719x_host_reset is there from initial commit so its better we wont remove this lock. I think we Patch v1 is correct fix(pass GFP_ATOMIC instead of GPF_KERNEL )
