> -----Original Message----- > From: Nils Holland [mailto:nholland at tisys.org] > Sent: Monday, January 23, 2017 7:31 PM > To: kbuild test robot > Cc: kbuild-all at 01.org; amd-gfx at lists.freedesktop.org; Deucher, Alexander; > Daenzer, Michel; Koenig, Christian > Subject: Re: [radeon-alex:drm-next-4.11-wip 131/131] > drivers/gpu/drm/amd/amdgpu/amdgpu_object.c:380:48: error: expected ')' > before ';' token > > On Tue, Jan 24, 2017 at 06:37:58AM +0800, kbuild test robot wrote: > > > > drivers/gpu/drm/amd/amdgpu/amdgpu_object.c: In function > 'amdgpu_bo_create_restricted': > > >> drivers/gpu/drm/amd/amdgpu/amdgpu_object.c:377:2: warning: > #warning Please enable CONFIG_MTRR and CONFIG_X86_PAT for better > performance thanks to write-combining [-Wcpp] > > #warning Please enable CONFIG_MTRR and CONFIG_X86_PAT for better > performance \ > > ^~~~~~~ > > >> drivers/gpu/drm/amd/amdgpu/amdgpu_object.c:380:48: error: > expected ')' before ';' token > > if (bo->flags & AMDGPU_GEM_CREATE_CPU_GTT_USWC;) > > ^ > > > > vim +380 drivers/gpu/drm/amd/amdgpu/amdgpu_object.c > > > > 371 #elif defined(CONFIG_X86) && !defined(CONFIG_X86_PAT) > > 372 /* Don't try to enable write-combining when it can't work, or > things > > 373 * may be slow > > 374 * See https://bugs.freedesktop.org/show_bug.cgi?id=88758 > > 375 */ > > 376 > > > 377 #warning Please enable CONFIG_MTRR and CONFIG_X86_PAT for > better performance \ > > 378 thanks to write-combining > > 379 > > > 380 if (bo->flags & AMDGPU_GEM_CREATE_CPU_GTT_USWC;) > > 381 DRM_INFO_ONCE("Please enable CONFIG_MTRR > and CONFIG_X86_PAT for " > > 382 "better performance thanks to write- > combining\n"); > > 383 bo->flags &= ~AMDGPU_GEM_CREATE_CPU_GTT_USWC; > > Whoops, the robot is certainly right, there's a ; that doesn't belong > in line 380. Shame on me! :-( > > Michel, now that you've already commited this patch, what do I do? > Should I submit another patch that just fixes this one issue, or > should I re-submit the whole patch, in fixed state, as a v2? > > Sorry for the hassle! :-/ I fixed this up locally when I committed it. It's already taken care of. Alex
