[PATCH 3/3] drm/amd/amdgpu: Various tidy ups for gfx6
StDenis, Tom
Tom.StDenis at amd.com
Fri Sep 2 12:08:21 UTC 2016
I'll submit a few new patches merging them in gfx 6/7/8 individually.
As for refactoring not really. A static check would only work if all of the code were in one unit. When I was adding the config data to debugfs I just grep'ed manually to discover that they're not really read anywhere in the kernel for instance.
Tom
________________________________
From: Christian König <deathsimple at vodafone.de>
Sent: Friday, September 2, 2016 08:06
To: StDenis, Tom; amd-gfx at lists.freedesktop.org
Subject: Re: [PATCH 3/3] drm/amd/amdgpu: Various tidy ups for gfx6
We could merge those to reduce the # of LOC.
Yes, please do so if you have time.
BTW: Do you know any tool/hack/whatever to find unused structure members? E.g. members which are never accessed or only set but never read?
We seem to have quite a number of those in amdgpu, radeon as well as TTM. Would be nice to clean such stuff up.
Christian.
Am 02.09.2016 um 13:54 schrieb StDenis, Tom:
Nope your eyes are fine. Note that the same pattern exists in gfx_v8_0.c as well. I guess different ring struct pointers are passed so they're not talking to the same ring.
We could merge those to reduce the # of LOC.
Tom
________________________________
From: Christian König <deathsimple at vodafone.de><mailto:deathsimple at vodafone.de>
Sent: Thursday, September 1, 2016 13:59
To: Tom St Denis; amd-gfx at lists.freedesktop.org<mailto:amd-gfx at lists.freedesktop.org>
Cc: StDenis, Tom
Subject: Re: [PATCH 3/3] drm/amd/amdgpu: Various tidy ups for gfx6
Am 01.09.2016 um 19:44 schrieb Tom St Denis:
> Various whitespace and logical simplifications for gfx6.
>
> Signed-off-by: Tom St Denis <tom.stdenis at amd.com><mailto:tom.stdenis at amd.com>
> ---
> drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c | 40 +++++------------------------------
> 1 file changed, 5 insertions(+), 35 deletions(-)
>
> diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c
> index 5f508c96496f..63ca77937714 100644
> --- a/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c
> +++ b/drivers/gpu/drm/amd/amdgpu/gfx_v6_0.c
> @@ -1211,11 +1211,8 @@ static void gfx_v6_0_gpu_init(struct amdgpu_device *adev)
> SC_EARLYZ_TILE_FIFO_SIZE(adev->gfx.config.sc_earlyz_tile_fifo_size)));
>
> WREG32(VGT_NUM_INSTANCES, 1);
> -
> WREG32(CP_PERFMON_CNTL, 0);
> -
> WREG32(SQ_CONFIG, 0);
> -
> WREG32(PA_SC_FORCE_EOV_MAX_CNTS, (FORCE_EOV_MAX_CLK_CNT(4095) |
> FORCE_EOV_MAX_REZ_CNT(255)));
>
> @@ -1240,7 +1237,6 @@ static void gfx_v6_0_gpu_init(struct amdgpu_device *adev)
> WREG32(PA_CL_ENHANCE, CLIP_VTX_REORDER_ENA | NUM_CLIP_SEQ(3));
>
> udelay(50);
> -
> }
>
>
> @@ -1661,21 +1657,14 @@ static int gfx_v6_0_cp_gfx_resume(struct amdgpu_device *adev)
>
> static u32 gfx_v6_0_ring_get_rptr_gfx(struct amdgpu_ring *ring)
> {
> - u32 rptr;
> -
> - rptr = ring->adev->wb.wb[ring->rptr_offs];
> -
> - return rptr;
> + return ring->adev->wb.wb[ring->rptr_offs];
> }
>
> static u32 gfx_v6_0_ring_get_wptr_gfx(struct amdgpu_ring *ring)
> {
> struct amdgpu_device *adev = ring->adev;
> - u32 wptr;
>
> - wptr = RREG32(CP_RB0_WPTR);
> -
> - return wptr;
> + return RREG32(CP_RB0_WPTR);
> }
>
> static void gfx_v6_0_ring_set_wptr_gfx(struct amdgpu_ring *ring)
> @@ -1688,9 +1677,7 @@ static void gfx_v6_0_ring_set_wptr_gfx(struct amdgpu_ring *ring)
>
> static u32 gfx_v6_0_ring_get_rptr_compute(struct amdgpu_ring *ring)
> {
> - u32 rptr = ring->adev->wb.wb[ring->rptr_offs];
> -
> - return rptr;
> + return ring->adev->wb.wb[ring->rptr_offs];
> }
Am I blind or are the gfx_v6_0_ring_get_rptr_compute() and
gfx_v6_0_ring_get_rptr_gfx() functions identical?
If that's true might be a good idea to just use one function.
Either way patch is Reviewed-by: Christian König <christian.koenig at amd.com><mailto:christian.koenig at amd.com>
Christian.
>
> static u32 gfx_v6_0_ring_get_wptr_compute(struct amdgpu_ring *ring)
> @@ -1770,14 +1757,12 @@ static int gfx_v6_0_cp_compute_resume(struct amdgpu_device *adev)
> ring->wptr = 0;
> WREG32(CP_RB1_WPTR, ring->wptr);
>
> -
> rptr_addr = adev->wb.gpu_addr + (ring->rptr_offs * 4);
> WREG32(CP_RB1_RPTR_ADDR, lower_32_bits(rptr_addr));
> WREG32(CP_RB1_RPTR_ADDR_HI, upper_32_bits(rptr_addr) & 0xFF);
>
> mdelay(1);
> WREG32(CP_RB1_CNTL, tmp);
> -
> WREG32(CP_RB1_BASE, ring->gpu_addr >> 8);
>
> ring = &adev->gfx.compute_ring[1];
> @@ -1797,7 +1782,6 @@ static int gfx_v6_0_cp_compute_resume(struct amdgpu_device *adev)
>
> mdelay(1);
> WREG32(CP_RB2_CNTL, tmp);
> -
> WREG32(CP_RB2_BASE, ring->gpu_addr >> 8);
>
> adev->gfx.compute_ring[0].ready = true;
> @@ -1825,12 +1809,7 @@ static void gfx_v6_0_cp_enable(struct amdgpu_device *adev, bool enable)
>
> static int gfx_v6_0_cp_load_microcode(struct amdgpu_device *adev)
> {
> - int r;
> -
> - r = gfx_v6_0_cp_gfx_load_microcode(adev);
> -
> - return r;
> -
> + return gfx_v6_0_cp_gfx_load_microcode(adev);
> }
>
> static void gfx_v6_0_enable_gui_idle_interrupt(struct amdgpu_device *adev,
> @@ -2172,7 +2151,6 @@ static void gfx_v6_0_rlc_stop(struct amdgpu_device *adev)
> WREG32(RLC_CNTL, 0);
>
> gfx_v6_0_enable_gui_idle_interrupt(adev, false);
> -
> gfx_v6_0_wait_for_rlc_serdes(adev);
> }
>
> @@ -2223,11 +2201,8 @@ static int gfx_v6_0_rlc_resume(struct amdgpu_device *adev)
> return -EINVAL;
>
> gfx_v6_0_rlc_stop(adev);
> -
> gfx_v6_0_rlc_reset(adev);
> -
> gfx_v6_0_init_pg(adev);
> -
> gfx_v6_0_init_cg(adev);
>
> WREG32(RLC_RL_BASE, 0);
> @@ -2254,7 +2229,6 @@ static int gfx_v6_0_rlc_resume(struct amdgpu_device *adev)
> WREG32(RLC_UCODE_ADDR, 0);
>
> gfx_v6_0_enable_lbpw(adev, gfx_v6_0_lbpw_supported(adev));
> -
> gfx_v6_0_rlc_start(adev);
>
> return 0;
> @@ -2278,7 +2252,6 @@ static void gfx_v6_0_enable_cgcg(struct amdgpu_device *adev, bool enable)
> WREG32(RLC_SERDES_WR_CTRL, 0x00b000ff);
>
> gfx_v6_0_wait_for_rlc_serdes(adev);
> -
> gfx_v6_0_update_rlc(adev, tmp);
>
> WREG32(RLC_SERDES_WR_CTRL, 0x007000ff);
> @@ -2931,13 +2904,10 @@ static bool gfx_v6_0_is_idle(void *handle)
> static int gfx_v6_0_wait_for_idle(void *handle)
> {
> unsigned i;
> - u32 tmp;
> struct amdgpu_device *adev = (struct amdgpu_device *)handle;
>
> for (i = 0; i < adev->usec_timeout; i++) {
> - tmp = RREG32(GRBM_STATUS) & GRBM_STATUS__GUI_ACTIVE_MASK;
> -
> - if (!tmp)
> + if (gfx_v6_0_is_idle(handle))
> return 0;
> udelay(1);
> }
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.freedesktop.org/archives/amd-gfx/attachments/20160902/d0eacb7a/attachment.html>
More information about the amd-gfx
mailing list