This affects hw_init only. It's not on the path of S3/resume and gpu_recovery/resume. So, they are not affected. Regards, Evan > -----Original Message----- > From: Xu, Feifei > Sent: 2019年3月22日 16:56 > To: Quan, Evan <Evan.Quan@xxxxxxx>; amd-gfx@xxxxxxxxxxxxxxxxxxxxx > Cc: Quan, Evan <Evan.Quan@xxxxxxx> > Subject: RE: [PATCH] drm/amdgpu: defer cmd/fence/fw buffers destroy on > hw_init failure > > Have you tested in S3 and gpu_recovery? > > If yes. Reviewed-by: Feifei Xu <Feifei.Xu@xxxxxxx> > > -----Original Message----- > From: amd-gfx <amd-gfx-bounces@xxxxxxxxxxxxxxxxxxxxx> On Behalf Of Evan > Quan > Sent: Monday, March 18, 2019 10:42 AM > To: amd-gfx@xxxxxxxxxxxxxxxxxxxxx > Cc: Quan, Evan <Evan.Quan@xxxxxxx> > Subject: [PATCH] drm/amdgpu: defer cmd/fence/fw buffers destroy on > hw_init failure > > As the cleanup jobs performed in pre_fini may still need these buffers. NULL > pointer dereference will be triggered without them. > > Change-Id: I398efd652584e022debf237950207199a4ea78fc > Signed-off-by: Evan Quan <evan.quan@xxxxxxx> > --- > drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c | 31 ++++++++++-------------- > - > 1 file changed, 12 insertions(+), 19 deletions(-) > > diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c > b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c > index 5888e24219d9..2206bb4b0903 100644 > --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c > +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c > @@ -890,59 +890,52 @@ static int psp_load_fw(struct amdgpu_device > *adev) > &psp->fence_buf_mc_addr, > &psp->fence_buf); > if (ret) > - goto failed_mem2; > + goto failed; > > ret = amdgpu_bo_create_kernel(adev, PSP_CMD_BUFFER_SIZE, > PAGE_SIZE, > AMDGPU_GEM_DOMAIN_VRAM, > &psp->cmd_buf_bo, &psp- > >cmd_buf_mc_addr, > (void **)&psp->cmd_buf_mem); > if (ret) > - goto failed_mem1; > + goto failed; > > memset(psp->fence_buf, 0, PSP_FENCE_BUFFER_SIZE); > > ret = psp_ring_init(psp, PSP_RING_TYPE__KM); > if (ret) { > DRM_ERROR("PSP ring init failed!\n"); > - goto failed_mem; > + goto failed; > } > > ret = psp_tmr_init(psp); > if (ret) { > DRM_ERROR("PSP tmr init failed!\n"); > - goto failed_mem; > + goto failed; > } > > ret = psp_asd_init(psp); > if (ret) { > DRM_ERROR("PSP asd init failed!\n"); > - goto failed_mem; > + goto failed; > } > > skip_memalloc: > ret = psp_hw_start(psp); > if (ret) > - goto failed_mem; > + goto failed; > > ret = psp_np_fw_load(psp); > if (ret) > - goto failed_mem; > + goto failed; > > return 0; > > -failed_mem: > - amdgpu_bo_free_kernel(&psp->cmd_buf_bo, > - &psp->cmd_buf_mc_addr, > - (void **)&psp->cmd_buf_mem); > -failed_mem1: > - amdgpu_bo_free_kernel(&psp->fence_buf_bo, > - &psp->fence_buf_mc_addr, &psp->fence_buf); > -failed_mem2: > - amdgpu_bo_free_kernel(&psp->fw_pri_bo, > - &psp->fw_pri_mc_addr, &psp->fw_pri_buf); > failed: > - kfree(psp->cmd); > - psp->cmd = NULL; > + /* > + * all cleanup jobs (xgmi terminate, ras terminate, > + * ring destroy, cmd/fence/fw buffers destory, > + * psp->cmd destory) are delayed to psp_hw_fini > + */ > return ret; > } > > -- > 2.21.0 > > _______________________________________________ > amd-gfx mailing list > amd-gfx@xxxxxxxxxxxxxxxxxxxxx > https://lists.freedesktop.org/mailman/listinfo/amd-gfx _______________________________________________ amd-gfx mailing list amd-gfx@xxxxxxxxxxxxxxxxxxxxx https://lists.freedesktop.org/mailman/listinfo/amd-gfx