From mboxrd@z Thu Jan 1 00:00:00 1970 From: =?windows-1252?Q?Michel_D=E4nzer?= Subject: Re: [PATCH] drm/radeon: Sync ME and PFP after CP semaphore waits v2 Date: Tue, 19 Aug 2014 14:57:11 +0900 Message-ID: <53F2E737.4020907@daenzer.net> References: <1408372212-18954-1-git-send-email-deathsimple@vodafone.de> Mime-Version: 1.0 Content-Type: text/plain; charset="windows-1252" Content-Transfer-Encoding: quoted-printable Return-path: Received: from mail.gna.ch (darkcity.gna.ch [195.226.6.51]) by gabe.freedesktop.org (Postfix) with ESMTP id E33A66E13D for ; Mon, 18 Aug 2014 22:57:18 -0700 (PDT) In-Reply-To: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: dri-devel-bounces@lists.freedesktop.org Sender: "dri-devel" To: Alex Deucher , =?windows-1252?Q?Christian_K=F6?= =?windows-1252?Q?nig?= Cc: Maling list - DRI developers List-Id: dri-devel@lists.freedesktop.org On 19.08.2014 00:02, Alex Deucher wrote: > On Mon, Aug 18, 2014 at 10:30 AM, Christian K=F6nig > wrote: >> From: Christian K=F6nig >> >> Fixes lockups due to CP read GPUVM faults when running piglit on Cape >> Verde. >> >> v2 (chk): apply the fix to R600+ as well, on CIK only the GFX CP has >> a PFP, add more comments to R600 code, enable flushing again >> >> Signed-off-by: Michel D=E4nzer >> Signed-off-by: Christian K=F6nig > = > I just checked the ucode. This packet only exists on 7xx and newer. > How about the attached change? Tested-by: Michel D=E4nzer Just survived 10 piglit runs in a row on my Cape Verde. :) >> diff --git a/drivers/gpu/drm/radeon/r600d.h b/drivers/gpu/drm/radeon/r60= 0d.h >> index f94e7a9..1272c0d 100644 >> --- a/drivers/gpu/drm/radeon/r600d.h >> +++ b/drivers/gpu/drm/radeon/r600d.h >> @@ -1597,6 +1597,7 @@ >> */ >> # define PACKET3_CP_DMA_CMD_SAIC (1 << 28) >> # define PACKET3_CP_DMA_CMD_DAIC (1 << 29) >> +#define PACKET3_PFP_SYNC_ME 0x42 >> #define PACKET3_SURFACE_SYNC 0x43 >> # define PACKET3_CB0_DEST_BASE_ENA (1 << 6) >> # define PACKET3_FULL_CACHE_ENA (1 << 20) /* r7xx+ o= nly */ The indentation doesn't seem to line up. >> diff --git a/drivers/gpu/drm/radeon/radeon_vm.c b/drivers/gpu/drm/radeon= /radeon_vm.c >> index 058f200..9c8358f 100644 >> --- a/drivers/gpu/drm/radeon/radeon_vm.c >> +++ b/drivers/gpu/drm/radeon/radeon_vm.c >> @@ -238,9 +238,7 @@ void radeon_vm_flush(struct radeon_device *rdev, >> uint64_t pd_addr =3D radeon_bo_gpu_offset(vm->page_directory); >> >> /* if we can't remember our last VM flush then flush now! */ >> - /* XXX figure out why we have to flush all the time before CIK */ >> - if (rdev->family < CHIP_BONAIRE || >> - !vm->last_flush || pd_addr !=3D vm->pd_gpu_addr) { >> + if (!vm->last_flush || pd_addr !=3D vm->pd_gpu_addr) { >> trace_radeon_vm_flush(pd_addr, ring, vm->id); >> vm->pd_gpu_addr =3D pd_addr; >> radeon_ring_vm_flush(rdev, ring, vm); Might be better for this to be in a separate change. -- = Earthling Michel D=E4nzer | http://www.amd.com Libre software enthusiast | Mesa and X developer