| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <d0a0ecbb-7bd6-4dfa-8fd2-dce3db317862@linaro.org>
Date: Thu, 25 Sep 2025 23:59:19 +0100
From: Bryan O'Donoghue <bryan.odonoghue@...aro.org>
To: Jingyi Wang <jingyi.wang@....qualcomm.com>,
Loic Poulain <loic.poulain@....qualcomm.com>, Robert Foss
<rfoss@...nel.org>, Andi Shyti <andi.shyti@...nel.org>,
Rob Herring <robh@...nel.org>, Krzysztof Kozlowski <krzk+dt@...nel.org>,
Conor Dooley <conor+dt@...nel.org>, Todor Tomov <todor.too@...il.com>,
Vladimir Zapolskiy <vladimir.zapolskiy@...aro.org>,
Mauro Carvalho Chehab <mchehab@...nel.org>,
Hangxiang Ma <hangxiang.ma@....qualcomm.com>,
Bryan O'Donoghue <bryan.odonoghue@...aro.org>
Cc: linux-i2c@...r.kernel.org, linux-arm-msm@...r.kernel.org,
devicetree@...r.kernel.org, linux-kernel@...r.kernel.org,
linux-media@...r.kernel.org, aiqun.yu@....qualcomm.com,
tingwei.zhang@....qualcomm.com, trilok.soni@....qualcomm.com,
yijie.yang@....qualcomm.com, Atiya Kailany <atiya.kailany@....qualcomm.com>
Subject: Re: [PATCH 6/6] media: qcom: camss: vfe: Add support for VFE 1080
On 25/09/2025 01:02, Jingyi Wang wrote:
> From: Hangxiang Ma <hangxiang.ma@....qualcomm.com>
>
> Add more detailed resource information for VFE and VFE Lite devices
> along with a driver for the 1080 version of those that is responsible
> for bus write master configuration and buffer address update.
"Add Video Front End (VFE) version 1080 as found on the Kaanapali SoC"
> The FULL front end modules in Kaanapali camera subsystem are called TFEs
> (Thin Front End), however, retaining the name VFE at places to maintain
> consistency and avoid unnecessary code changes.
Good
>
> Co-developed-by: Atiya Kailany <atiya.kailany@....qualcomm.com>
> Signed-off-by: Atiya Kailany <atiya.kailany@....qualcomm.com>
> Signed-off-by: Hangxiang Ma <hangxiang.ma@....qualcomm.com>
> Signed-off-by: Jingyi Wang <jingyi.wang@....qualcomm.com>
> ---
> drivers/media/platform/qcom/camss/Makefile | 1 +
> drivers/media/platform/qcom/camss/camss-vfe-1080.c | 156 +++++++++++++++++++++
> drivers/media/platform/qcom/camss/camss-vfe.c | 15 +-
> drivers/media/platform/qcom/camss/camss-vfe.h | 1 +
> drivers/media/platform/qcom/camss/camss.c | 138 ++++++++++++++++++
> 5 files changed, 307 insertions(+), 4 deletions(-)
>
> diff --git a/drivers/media/platform/qcom/camss/Makefile b/drivers/media/platform/qcom/camss/Makefile
> index 3a7ed4f5a004..dc41b0d6dc21 100644
> --- a/drivers/media/platform/qcom/camss/Makefile
> +++ b/drivers/media/platform/qcom/camss/Makefile
> @@ -22,6 +22,7 @@ qcom-camss-objs += \
> camss-vfe-340.o \
> camss-vfe-480.o \
> camss-vfe-680.o \
> + camss-vfe-1080.o \
> camss-vfe-gen3.o \
> camss-vfe-gen1.o \
> camss-vfe.o \
> diff --git a/drivers/media/platform/qcom/camss/camss-vfe-1080.c b/drivers/media/platform/qcom/camss/camss-vfe-1080.c
> new file mode 100644
> index 000000000000..f1852a3bf97f
> --- /dev/null
> +++ b/drivers/media/platform/qcom/camss/camss-vfe-1080.c
> @@ -0,0 +1,156 @@
> +// SPDX-License-Identifier: GPL-2.0
> +/*
> + * camss-vfe-1080.c
> + *
> + * Qualcomm MSM Camera Subsystem - VFE (Video Front End) Module v1080
> + *
> + * Copyright (c) Qualcomm Technologies, Inc. and/or its subsidiaries.
> + */
> +#include <linux/interrupt.h>
> +#include <linux/io.h>
> +#include <linux/iopoll.h>
> +
> +#include "camss.h"
> +#include "camss-vfe.h"
> +
> +/* VFE-1080 Bus Register Base Addresses */
> +#define BUS_REG_BASE (vfe_is_lite(vfe) ? 0x800 : 0x1000)
> +
> +#define VFE_BUS_WM_CGC_OVERRIDE (BUS_REG_BASE + 0x08)
> +#define WM_CGC_OVERRIDE_ALL (0x7FFFFFF)
> +
> +#define VFE_BUS_WM_TEST_BUS_CTRL (BUS_REG_BASE + 0x128)
> +
> +#define VFE_BUS_WM_CFG(n) (BUS_REG_BASE + 0x500 + (n) * 0x100)
> +#define WM_CFG_EN BIT(0)
> +#define WM_VIR_FRM_EN BIT(1)
> +#define WM_CFG_MODE BIT(16)
> +#define VFE_BUS_WM_IMAGE_ADDR(n) (BUS_REG_BASE + 0x504 + (n) * 0x100)
> +#define VFE_BUS_WM_FRAME_INCR(n) (BUS_REG_BASE + 0x508 + (n) * 0x100)
> +#define VFE_BUS_WM_IMAGE_CFG_0(n) (BUS_REG_BASE + 0x50c + (n) * 0x100)
> +#define WM_IMAGE_CFG_0_DEFAULT_WIDTH (0xFFFF)
> +#define VFE_BUS_WM_IMAGE_CFG_2(n) (BUS_REG_BASE + 0x514 + (n) * 0x100)
> +#define WM_IMAGE_CFG_2_DEFAULT_STRIDE (0xFFFF)
> +#define VFE_BUS_WM_PACKER_CFG(n) (BUS_REG_BASE + 0x518 + (n) * 0x100)
> +
> +#define VFE_BUS_WM_IRQ_SUBSAMPLE_PERIOD(n) (BUS_REG_BASE + 0x530 + (n) * 0x100)
> +#define VFE_BUS_WM_IRQ_SUBSAMPLE_PATTERN(n) (BUS_REG_BASE + 0x534 + (n) * 0x100)
> +
> +/* VFE lite has no such registers */
> +#define VFE_BUS_WM_FRAMEDROP_PERIOD(n) (BUS_REG_BASE + 0x538 + (n) * 0x100)
> +#define VFE_BUS_WM_FRAMEDROP_PATTERN(n) (BUS_REG_BASE + 0x53c + (n) * 0x100)
> +
> +#define VFE_BUS_WM_MMU_PREFETCH_CFG(n) (BUS_REG_BASE + 0x560 + (n) * 0x100)
> +#define VFE_BUS_WM_MMU_PREFETCH_MAX_OFFSET(n) (BUS_REG_BASE + 0x564 + (n) * 0x100)
> +
> +/*
> + * Bus client mapping:
> + *
> + * Full VFE:
> + * 23 = RDI0, 24 = RDI1, 25 = RDI2, 26 = RDI3, 27 = RDI4
> + *
> + * VFE LITE:
> + * 0 = RDI0, 1 = RDI1, 2 = RDI2, 3 = RDI3
> + */
Give the full bus client mapping, no the partial.
> +#define RDI_WM(n) ((vfe_is_lite(vfe) ? 0x0 : 0x17) + (n))
> +
> +static void vfe_wm_start_1080(struct vfe_device *vfe, u8 wm, struct vfe_line *line)
> +{
> + struct v4l2_pix_format_mplane *pix =
> + &line->video_out.active_fmt.fmt.pix_mp;
> +
> + wm = RDI_WM(wm);
> +
> + /* no clock gating at bus input */
> + writel(WM_CGC_OVERRIDE_ALL, vfe->base + VFE_BUS_WM_CGC_OVERRIDE);
> +
> + writel(0x0, vfe->base + VFE_BUS_WM_TEST_BUS_CTRL);
> +
> + writel(ALIGN(pix->plane_fmt[0].bytesperline, 16) * pix->height >> 8,
> + vfe->base + VFE_BUS_WM_FRAME_INCR(wm));
> + writel((WM_IMAGE_CFG_0_DEFAULT_WIDTH & 0xFFFF),
> + vfe->base + VFE_BUS_WM_IMAGE_CFG_0(wm));
> + writel(WM_IMAGE_CFG_2_DEFAULT_STRIDE,
> + vfe->base + VFE_BUS_WM_IMAGE_CFG_2(wm));
> + writel(0, vfe->base + VFE_BUS_WM_PACKER_CFG(wm));
> +
> + /* no dropped frames, one irq per frame */
> + if (!vfe_is_lite(vfe)) {
> + writel(0, vfe->base + VFE_BUS_WM_FRAMEDROP_PERIOD(wm));
> + writel(1, vfe->base + VFE_BUS_WM_FRAMEDROP_PATTERN(wm));
> + }
> +
> + writel(0, vfe->base + VFE_BUS_WM_IRQ_SUBSAMPLE_PERIOD(wm));
> + writel(1, vfe->base + VFE_BUS_WM_IRQ_SUBSAMPLE_PATTERN(wm));
> +
> + writel(1, vfe->base + VFE_BUS_WM_MMU_PREFETCH_CFG(wm));
> + writel(0xFFFFFFFF, vfe->base + VFE_BUS_WM_MMU_PREFETCH_MAX_OFFSET(wm));
> +
> + writel(WM_CFG_EN | WM_CFG_MODE, vfe->base + VFE_BUS_WM_CFG(wm));
> +}
> +
> +static void vfe_wm_stop_1080(struct vfe_device *vfe, u8 wm)
> +{
> + wm = RDI_WM(wm);
> + writel(0, vfe->base + VFE_BUS_WM_CFG(wm));
> +}
> +
> +static void vfe_wm_update_1080(struct vfe_device *vfe, u8 wm, u32 addr,
> + struct vfe_line *line)
> +{
> + wm = RDI_WM(wm);
> + writel((addr >> 8) & 0xFFFFFFFF, vfe->base + VFE_BUS_WM_IMAGE_ADDR(wm));
> +
> + dev_dbg(vfe->camss->dev, "wm:%d, image buf addr:0x%x\n", wm, addr);
> +}
> +
> +static void vfe_reg_update_1080(struct vfe_device *vfe, enum vfe_line_id line_id)
> +{
> + int port_id = line_id;
> +
> + camss_reg_update(vfe->camss, vfe->id, port_id, false);
> +}
> +
> +static const struct camss_video_ops vfe_video_ops_1080 = {
> + .queue_buffer = vfe_queue_buffer_v2,
> + .flush_buffers = vfe_flush_buffers,
> +};
> +
> +static void vfe_subdev_init_1080(struct device *dev, struct vfe_device *vfe)
> +{
> + vfe->video_ops = vfe_video_ops_1080;
> +}
> +
> +static void vfe_global_reset_1080(struct vfe_device *vfe)
> +{
> + vfe_isr_reset_ack(vfe);
> +}
> +
> +static irqreturn_t vfe_isr_1080(int irq, void *dev)
> +{
> + /* nop */
> + return IRQ_HANDLED;
> +}
> +
> +static int vfe_halt_1080(struct vfe_device *vfe)
> +{
> + /* rely on vfe_disable_output() to stop the VFE */
> + return 0;
> +}
> +
> +const struct vfe_hw_ops vfe_ops_1080 = {
> + .global_reset = vfe_global_reset_1080,
> + .hw_version = vfe_hw_version,
> + .isr = vfe_isr_1080,
> + .pm_domain_off = vfe_pm_domain_off,
> + .pm_domain_on = vfe_pm_domain_on,
> + .reg_update = vfe_reg_update_1080,
> + .subdev_init = vfe_subdev_init_1080,
> + .vfe_disable = vfe_disable,
> + .vfe_enable = vfe_enable_v2,
> + .vfe_halt = vfe_halt_1080,
> + .vfe_wm_start = vfe_wm_start_1080,
> + .vfe_wm_stop = vfe_wm_stop_1080,
> + .vfe_buf_done = vfe_buf_done,
> + .vfe_wm_update = vfe_wm_update_1080,
> +};
> diff --git a/drivers/media/platform/qcom/camss/camss-vfe.c b/drivers/media/platform/qcom/camss/camss-vfe.c
> index ee08dbbddf88..90e7fbd2428b 100644
> --- a/drivers/media/platform/qcom/camss/camss-vfe.c
> +++ b/drivers/media/platform/qcom/camss/camss-vfe.c
> @@ -349,6 +349,7 @@ static u32 vfe_src_pad_code(struct vfe_line *line, u32 sink_code,
> case CAMSS_845:
> case CAMSS_8550:
> case CAMSS_8775P:
> + case CAMSS_KAANAPALI:
> case CAMSS_X1E80100:
> switch (sink_code) {
> case MEDIA_BUS_FMT_YUYV8_1X16:
> @@ -521,7 +522,8 @@ int vfe_enable_output_v2(struct vfe_line *line)
>
> spin_lock_irqsave(&vfe->output_lock, flags);
>
> - ops->reg_update_clear(vfe, line->id);
> + if (ops->reg_update_clear)
> + ops->reg_update_clear(vfe, line->id);
What's going on here, why don't you have to clear the reg_update for
your hardware ?
>
> if (output->state > VFE_OUTPUT_RESERVED) {
> dev_err(vfe->camss->dev,
> @@ -541,14 +543,17 @@ int vfe_enable_output_v2(struct vfe_line *line)
>
> ops->vfe_wm_start(vfe, output->wm_idx[0], line);
>
> - for (i = 0; i < 2; i++) {
> + for (i = 0; i < CAMSS_INIT_BUF_COUNT; i++) {
This I like but it's not specific to VFE 1080.
It can/should be introduced as its own patch to remove this hard-coded
value.
> output->buf[i] = vfe_buf_get_pending(output);
> if (!output->buf[i])
> break;
> output->gen2.active_num++;
> ops->vfe_wm_update(vfe, output->wm_idx[0],
> output->buf[i]->addr[0], line);
> - ops->reg_update(vfe, line->id);
> +
> + /* Deferring the reg update until after CSID config */
> + if (vfe->camss->res->version != CAMSS_KAANAPALI)
> + ops->reg_update(vfe, line->id);
Needs more justification - here in the email thread will do.
OK you don't have to do reg_update but, your logic is not consistent for
this.
In one case you check for ops->reg_update_clear() in this case you check
for !KANNAPALI
Definitely don't want to have SoC specific logic in the core files
unless absolutely necessary, which in this case it is not.
> }
>
> spin_unlock_irqrestore(&vfe->output_lock, flags);
> @@ -914,7 +919,8 @@ static int vfe_match_clock_names(struct vfe_device *vfe,
> return (!strcmp(clock->name, vfe_name) ||
> !strcmp(clock->name, vfe_lite_name) ||
> !strcmp(clock->name, "vfe_lite") ||
> - !strcmp(clock->name, "camnoc_axi"));
> + !strcmp(clock->name, "camnoc_axi") ||
> + !strcmp(clock->name, "camnoc_rt_axi"));
I'd prefer to see a small patch adding camnoc_rt_axi and explaining what
it is inside of CAMSS and why we should clock it.
> }
>
> /*
> @@ -1997,6 +2003,7 @@ static int vfe_bpl_align(struct vfe_device *vfe)
> case CAMSS_845:
> case CAMSS_8550:
> case CAMSS_8775P:
> + case CAMSS_KAANAPALI:
> case CAMSS_X1E80100:
> ret = 16;
> break;
> diff --git a/drivers/media/platform/qcom/camss/camss-vfe.h b/drivers/media/platform/qcom/camss/camss-vfe.h
> index 0300efdb1c46..444924ddf724 100644
> --- a/drivers/media/platform/qcom/camss/camss-vfe.h
> +++ b/drivers/media/platform/qcom/camss/camss-vfe.h
> @@ -245,6 +245,7 @@ extern const struct vfe_hw_ops vfe_ops_170;
> extern const struct vfe_hw_ops vfe_ops_340;
> extern const struct vfe_hw_ops vfe_ops_480;
> extern const struct vfe_hw_ops vfe_ops_680;
> +extern const struct vfe_hw_ops vfe_ops_1080;
> extern const struct vfe_hw_ops vfe_ops_gen3;
>
> int vfe_get(struct vfe_device *vfe);
> diff --git a/drivers/media/platform/qcom/camss/camss.c b/drivers/media/platform/qcom/camss/camss.c
> index 74a8ad3cb728..2239abfef26e 100644
> --- a/drivers/media/platform/qcom/camss/camss.c
> +++ b/drivers/media/platform/qcom/camss/camss.c
> @@ -217,6 +217,142 @@ static const struct camss_subdev_resources csid_res_kaanapali[] = {
> }
> };
>
> +/* In Kaanapali, CAMNOC requires all CAMNOC_RT_TFEX clocks
> + * to operate on any TFE Full.
> + */
> +static const struct camss_subdev_resources vfe_res_kaanapali[] = {
> + /* VFE0 - TFE Full */
> + {
> + .regulators = {},
> + .clock = { "gcc_hf_axi", "vfe0_fast_ahb", "vfe0",
> + "camnoc_rt_vfe0", "camnoc_rt_vfe1", "camnoc_rt_vfe2",
> + "camnoc_rt_axi", "camnoc_nrt_axi", "qdss_debug_xo" },
XO debug ?
Really ? Why ?
Whats that clock for and why is it needed for the runtime TFE to function ?
---
bod
Powered by blists - more mailing lists