[AMD Official Use Only - General] Then please update commit message, this change is due to "value r is never used, and remove unnecessary assignment", that makes sense to me.
Thanks Ruijing -----Original Message----- From: Khalid Masum <[email protected]> Sent: Monday, August 15, 2022 11:54 AM To: Dong, Ruijing <[email protected]>; [email protected]; [email protected]; [email protected]; [email protected] Cc: Deucher, Alexander <[email protected]>; Koenig, Christian <[email protected]>; Pan, Xinhui <[email protected]>; David Airlie <[email protected]>; Daniel Vetter <[email protected]>; Zhu, James <[email protected]>; Jiang, Sonny <[email protected]>; Wan Jiabing <[email protected]>; Liu, Leo <[email protected]> Subject: Re: [PATCH linux-next] drm/amdgpu/vcn: Remove unused assignment in vcn_v4_0_stop On 8/15/22 21:17, Dong, Ruijing wrote: > [AMD Official Use Only - General] > > If the condition was met and it came to execute vcn_4_0_stop_dpg_mode, then > it would never have a chance to go for /*wait for vcn idle*/, isn't it? Hypothetically, some other thread might set adev->pg_flags NULL and in that case it will get the chance to go for /* wait for vcn idle */. > I still didn't see obvious purpose of this change. > > if (adev->pg_flags & AMD_PG_SUPPORT_VCN_DPG) { > ==> r = vcn_v4_0_stop_dpg_mode(adev, i); Regardless of that, this assignment to r is unnecessary. Because this value of r is never used. This patch simply removes this unnecessary assignment. > continue; > } > > /* wait for vcn idle */ > r = SOC15_WAIT_ON_RREG(VCN, i, regUVD_STATUS, > UVD_STATUS__IDLE, 0x7); > > Thanks > Ruijing > Thanks, -- Khalid Masum
