[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <20200526164057.GC20960@jcrouse1-lnx.qualcomm.com>
Date: Tue, 26 May 2020 10:40:57 -0600
From: Jordan Crouse <jcrouse@...eaurora.org>
To: Jonathan Marek <jonathan@...ek.ca>
Cc: freedreno@...ts.freedesktop.org, Rob Clark <robdclark@...il.com>,
Sean Paul <sean@...rly.run>, David Airlie <airlied@...ux.ie>,
Daniel Vetter <daniel@...ll.ch>,
Sharat Masetty <smasetty@...eaurora.org>,
Akhil P Oommen <akhilpo@...eaurora.org>,
"open list:DRM DRIVER FOR MSM ADRENO GPU"
<linux-arm-msm@...r.kernel.org>,
"open list:DRM DRIVER FOR MSM ADRENO GPU"
<dri-devel@...ts.freedesktop.org>,
open list <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] drm/msm/a6xx: set ubwc config for A640 and A650
On Mon, May 25, 2020 at 11:25:13PM -0400, Jonathan Marek wrote:
> This is required for A640 and A650 to be able to share UBWC-compressed
> images with other HW such as display, which expect this configuration.
> Signed-off-by: Jonathan Marek <jonathan@...ek.ca>
> ---
> drivers/gpu/drm/msm/adreno/a6xx_gpu.c | 38 ++++++++++++++++++++++-----
> 1 file changed, 32 insertions(+), 6 deletions(-)
>
> diff --git a/drivers/gpu/drm/msm/adreno/a6xx_gpu.c b/drivers/gpu/drm/msm/adreno/a6xx_gpu.c
> index 6f335ae179c8..aa004a261277 100644
> --- a/drivers/gpu/drm/msm/adreno/a6xx_gpu.c
> +++ b/drivers/gpu/drm/msm/adreno/a6xx_gpu.c
> @@ -289,6 +289,37 @@ static void a6xx_set_hwcg(struct msm_gpu *gpu, bool state)
> gpu_write(gpu, REG_A6XX_RBBM_CLOCK_CNTL, state ? 0x8aa8aa02 : 0);
> }
>
> +static void a6xx_set_ubwc_config(struct msm_gpu *gpu)
> +{
> + struct adreno_gpu *adreno_gpu = to_adreno_gpu(gpu);
> + u32 lower_bit = 2;
> + u32 amsbc = 0;
> + u32 rgb565_predicator = 0;
> + u32 uavflagprd_inv = 0;
> +
This hardware design has the amazing ability to make me sad every time I see it.
> + /* a618 is using the hw default values */
> + if (adreno_is_a618(adreno_gpu))
> + return;
> +
I've been a recent convert to the cult of <linux/bitfields.h> and FIELD_PREP()
could help, maybe?
if (adreno_is_a640(adreno_gpu)) {
rb_ncmode = FIELD_PREP(REG_A6XX_RB_NC_MODE_CNTL_AMSBC, 1);
rb_ncmode |= FIELD_PREP(REG_A6XX_RB_NC_MODE_CNTL_HBB, 2);
tpl1_ncmode = FIELD_PREP(REG_A6XX_TPL1_NC_MODE_HBB, 2);
sp_ncmode = FIELD_PREP(REG_A6XX_TPL1_NC_MODE_HBB, 2);
uchemode = FIELD_PREP(REG_A6XX_UCHE_MODE_CNTL_HBB, 2);
} else if adreno_is_a650(adreno_gpu)) {
rb_ncmode = FIELD_PREP(REG_A6XX_RB_NC_MODE_CNTL_AMSBC, 1);
rb_ncmode |= FIELD_PREP(REG_A6XX_RB_NC_MODE_CNTL_HBB, 3);
rb_ncmode |= FIELD_PREP(REG_A6XX_RB_NC_MODE_CNTL_RGB565, 1);
tpl1_ncmode = FIELD_PREP(REG_A6XX_TPL1_NC_MODE_HBB, 3);
sp_ncmode = FIELD_PREP(REG_A6XX_TPL1_NC_MODE_HBB, 3);
sp_ncmode |= FIELD_PREP(REG_A6XX_TPL1_NC_MODE_UAVFLAGPRD_INV, 2);
uchemode = FIELD_PREP(REG_A6XX_UCHE_MODE_CNTL_HBB, 2);
}
I'm not sure if that is any clearer or not. Perhaps this is a problem for the
next person to add a new target. Regardless the code is programming the
hardware correctly so...
Reviewed-by: Jordan Crouse <jcrouse@...eaurora.org>
> + if (adreno_is_a640(adreno_gpu))
> + amsbc = 1;
> +
> + if (adreno_is_a650(adreno_gpu)) {
> + /* TODO: get ddr type from bootloader and use 2 for LPDDR4 */
> + lower_bit = 3;
> + amsbc = 1;
> + rgb565_predicator = 1;
> + uavflagprd_inv = 2;
> + }
> +
> + gpu_write(gpu, REG_A6XX_RB_NC_MODE_CNTL,
> + rgb565_predicator << 11 | amsbc << 4 | lower_bit << 1);
> + gpu_write(gpu, REG_A6XX_TPL1_NC_MODE_CNTL, lower_bit << 1);
> + gpu_write(gpu, REG_A6XX_SP_NC_MODE_CNTL,
> + uavflagprd_inv >> 4 | lower_bit << 1);
> + gpu_write(gpu, REG_A6XX_UCHE_MODE_CNTL, lower_bit << 21);
> +}
> +
> static int a6xx_cp_init(struct msm_gpu *gpu)
> {
> struct msm_ringbuffer *ring = gpu->rb[0];
> @@ -478,12 +509,7 @@ static int a6xx_hw_init(struct msm_gpu *gpu)
> /* Select CP0 to always count cycles */
> gpu_write(gpu, REG_A6XX_CP_PERFCTR_CP_SEL_0, PERF_CP_ALWAYS_COUNT);
>
> - if (adreno_is_a630(adreno_gpu)) {
> - gpu_write(gpu, REG_A6XX_RB_NC_MODE_CNTL, 2 << 1);
> - gpu_write(gpu, REG_A6XX_TPL1_NC_MODE_CNTL, 2 << 1);
> - gpu_write(gpu, REG_A6XX_SP_NC_MODE_CNTL, 2 << 1);
> - gpu_write(gpu, REG_A6XX_UCHE_MODE_CNTL, 2 << 21);
> - }
> + a6xx_set_ubwc_config(gpu);
>
> /* Enable fault detection */
> gpu_write(gpu, REG_A6XX_RBBM_INTERFACE_HANG_INT_CNTL,
> --
> 2.26.1
>
--
The Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum,
a Linux Foundation Collaborative Project
Powered by blists - more mailing lists