| 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: <6ab0ab15-f78f-5f68-e699-e75f1dbb993e@quicinc.com>
Date: Wed, 20 Dec 2023 13:33:43 +0530
From: Dikshita Agarwal <quic_dikshita@...cinc.com>
To: Dmitry Baryshkov <dmitry.baryshkov@...aro.org>,
<linux-media@...r.kernel.org>, <linux-kernel@...r.kernel.org>,
<stanimir.k.varbanov@...il.com>, <quic_vgarodia@...cinc.com>,
<agross@...nel.org>, <andersson@...nel.org>,
<konrad.dybcio@...aro.org>, <mchehab@...nel.org>,
<bryan.odonoghue@...aro.org>
CC: <linux-arm-msm@...r.kernel.org>, <quic_abhinavk@...cinc.com>
Subject: Re: [PATCH v2 02/34] media: introduce common helpers for queues
handling
On 12/18/2023 11:59 PM, Dmitry Baryshkov wrote:
> On 18/12/2023 13:31, Dikshita Agarwal wrote:
>> Introduce common helpers for shared queues init, deinit
>> and read/write queue operations and use them in venus driver.
>>
>> Signed-off-by: Dikshita Agarwal <quic_dikshita@...cinc.com>
>> ---
>> drivers/media/platform/qcom/vcodec/hfi_queue.c | 258 +++++++++++
>> drivers/media/platform/qcom/vcodec/hfi_queue.h | 129 ++++++
>> drivers/media/platform/qcom/vcodec/venus/Makefile | 3 +-
>> .../media/platform/qcom/vcodec/venus/hfi_venus.c | 500
>> ++++-----------------
>> 4 files changed, 465 insertions(+), 425 deletions(-)
>> create mode 100644 drivers/media/platform/qcom/vcodec/hfi_queue.c
>> create mode 100644 drivers/media/platform/qcom/vcodec/hfi_queue.h
>>
>> diff --git a/drivers/media/platform/qcom/vcodec/hfi_queue.c
>> b/drivers/media/platform/qcom/vcodec/hfi_queue.c
>> new file mode 100644
>> index 0000000..e3e470b
>> --- /dev/null
>> +++ b/drivers/media/platform/qcom/vcodec/hfi_queue.c
>> @@ -0,0 +1,258 @@
>> +// SPDX-License-Identifier: GPL-2.0-only
>> +/*
>> + * Copyright (c) 2022-2023 Qualcomm Innovation Center, Inc. All rights
>> reserved.
>> + */
>> +
>> +#include "hfi_queue.h"
>> +
>> +int write_queue(struct iface_q_info *qinfo, void *packet,
>> + u32 packet_size, u32 *rx_req)
>
> Same issue here: no prefixes.
>
Sure, will fix the API names with proper prefixes.
>> +{
>> + u32 empty_space, read_idx, write_idx, new_write_idx;
>> + struct hfi_queue_header *queue;
>> + u32 *write_ptr, residue;
>> +
>> + queue = qinfo->qhdr;
>> + if (!queue)
>> + return -EINVAL;
>> +
>> + if (!packet_size || packet_size > qinfo->q_array.size)
>> + return -EINVAL;
>> +
>> + read_idx = queue->qhdr_read_idx * sizeof(u32);
>> + write_idx = queue->qhdr_write_idx * sizeof(u32);
>> +
>> + empty_space = (write_idx >= read_idx) ?
>> + (qinfo->q_array.size - (write_idx - read_idx)) :
>> + (read_idx - write_idx);
>> + if (empty_space <= packet_size) {
>> + queue->qhdr_tx_req = 1;
>> + return -ENOSPC;
>> + }
>> +
>> + queue->qhdr_tx_req = 0;
>> +
>> + new_write_idx = write_idx + packet_size;
>> + write_ptr = (u32 *)((u8 *)qinfo->q_array.kernel_vaddr + write_idx);
>> +
>> + if (write_ptr < (u32 *)qinfo->q_array.kernel_vaddr ||
>> + write_ptr > (u32 *)(qinfo->q_array.kernel_vaddr +
>> + qinfo->q_array.size))
>> + return -EINVAL;
>> +
>> + if (new_write_idx < qinfo->q_array.size) {
>> + memcpy(write_ptr, packet, packet_size);
>> + } else {
>> + residue = new_write_idx - qinfo->q_array.size;
>> + memcpy(write_ptr, packet, (packet_size - residue));
>> + memcpy(qinfo->q_array.kernel_vaddr,
>> + packet + (packet_size - residue), residue);
>> + new_write_idx = residue;
>> + }
>> +
>> + /* Make sure packet is written before updating the write index */
>> + mb();
>> + queue->qhdr_write_idx = new_write_idx / sizeof(u32);
>> + *rx_req = queue->qhdr_rx_req ? 1 : 0;
>> +
>> + /* Make sure write index is updated before an interrupt is raised */
>> + mb();
>> +
>> + return 0;
>> +}
>> +
>> +int read_queue(struct iface_q_info *qinfo, void *packet, u32 *tx_req)
>> +{
>> + u32 read_idx, write_idx, new_read_idx;
>> + struct hfi_queue_header *queue;
>> + u32 receive_request = 0;
>> + u32 packet_size, residue;
>> + u32 *read_ptr;
>> + int ret = 0;
>> +
>> + /* Make sure data is valid before reading it */
>> + mb();
>> + queue = qinfo->qhdr;
>> + if (!queue)
>> + return -EINVAL;
>> +
>> + if (queue->qhdr_type & IFACEQ_MSGQ_ID)
>> + receive_request = 1;
>> +
>> + read_idx = queue->qhdr_read_idx * sizeof(u32);
>> + write_idx = queue->qhdr_write_idx * sizeof(u32);
>> +
>> + if (read_idx == write_idx) {
>> + queue->qhdr_rx_req = receive_request;
>> + /* Ensure qhdr is updated in main memory */
>> + mb();
>> + return -ENODATA;
>> + }
>> +
>> + read_ptr = (u32 *)(qinfo->q_array.kernel_vaddr + read_idx);
>> + if (read_ptr < (u32 *)qinfo->q_array.kernel_vaddr ||
>> + read_ptr > (u32 *)(qinfo->q_array.kernel_vaddr +
>> + qinfo->q_array.size - sizeof(*read_ptr)))
>> + return -ENODATA;
>> +
>> + packet_size = *read_ptr;
>> + if (!packet_size)
>> + return -EINVAL;
>> +
>> + new_read_idx = read_idx + packet_size;
>> + if (packet_size <= IFACEQ_CORE_PKT_SIZE &&
>> + read_idx <= qinfo->q_array.size) {
>> + if (new_read_idx < qinfo->q_array.size) {
>> + memcpy(packet, read_ptr, packet_size);
>> + } else {
>> + residue = new_read_idx - qinfo->q_array.size;
>> + memcpy(packet, read_ptr, (packet_size - residue));
>> + memcpy((packet + (packet_size - residue)),
>> + qinfo->q_array.kernel_vaddr, residue);
>> + new_read_idx = residue;
>> + }
>> + } else {
>> + new_read_idx = write_idx;
>> + ret = -EBADMSG;
>> + }
>> +
>> + queue->qhdr_rx_req = receive_request;
>> +
>> + queue->qhdr_read_idx = new_read_idx / sizeof(u32);
>> +
>> + *tx_req = queue->qhdr_tx_req ? 1 : 0;
>> + /* Ensure qhdr is updated in main memory */
>> + mb();
>> +
>> + return ret;
>> +}
>> +
>> +static void __set_queue_hdr_defaults(struct hfi_queue_header *q_hdr)
>> +{
>> + q_hdr->qhdr_status = 0x1;
>> + q_hdr->qhdr_type = IFACEQ_DFLT_QHDR;
>> + q_hdr->qhdr_q_size = IFACEQ_QUEUE_SIZE / 4;
>> + q_hdr->qhdr_pkt_size = 0;
>> + q_hdr->qhdr_rx_wm = 0x1;
>> + q_hdr->qhdr_tx_wm = 0x1;
>> + q_hdr->qhdr_rx_req = 0x1;
>> + q_hdr->qhdr_tx_req = 0x0;
>> + q_hdr->qhdr_rx_irq_status = 0x0;
>> + q_hdr->qhdr_tx_irq_status = 0x0;
>> + q_hdr->qhdr_read_idx = 0x0;
>> + q_hdr->qhdr_write_idx = 0x0;
>> +}
>> +
>> +static void hfi_set_queue_header(u32 queue_id, struct iface_q_info
>> *iface_q)
>> +{
>> + __set_queue_hdr_defaults(iface_q->qhdr);
>> + iface_q->qhdr->qhdr_start_addr = iface_q->q_array.device_addr;
>> + iface_q->qhdr->qhdr_type |= queue_id;
>> +
>> + /*
>> + * Set receive request to zero on debug queue as there is no
>> + * need of interrupt from video hardware for debug messages
>> + */
>> + if (queue_id == IFACEQ_DBGQ_ID)
>> + iface_q->qhdr->qhdr_rx_req = 0;
>> +}
>> +
>> +static void __queue_init(struct mem_desc *q_table, u32 queue_id, struct
>> iface_q_info *iface_q)
>> +{
>> + unsigned int offset = 0;
>> +
>> + offset = q_table->size + (queue_id * IFACEQ_QUEUE_SIZE);
>> + iface_q->q_array.device_addr = q_table->device_addr + offset;
>> + iface_q->q_array.kernel_vaddr =
>> + (void *)((char *)q_table->kernel_vaddr + offset);
>> + iface_q->q_array.size = IFACEQ_QUEUE_SIZE;
>> + iface_q->qhdr =
>> + IFACEQ_GET_QHDR_START_ADDR(q_table->kernel_vaddr, queue_id);
>> +
>> + memset(iface_q->qhdr, 0, sizeof(struct hfi_queue_header));
>> +
>> + hfi_set_queue_header(queue_id, iface_q);
>> +}
>> +
>> +int hfi_queue_init(struct device *dev, struct mem_desc *q_table, struct
>> mem_desc *sfr,
>> + struct iface_q_info *cmd_q, struct iface_q_info *msg_q,
>> + struct iface_q_info *dbg_q, void *daddr)
>> +{
>> + struct hfi_queue_table_header *q_tbl_hdr;
>> +
>> + if (q_table->kernel_vaddr) {
>> + hfi_set_queue_header(IFACEQ_CMDQ_ID, cmd_q);
>> + hfi_set_queue_header(IFACEQ_MSGQ_ID, msg_q);
>> + hfi_set_queue_header(IFACEQ_DBGQ_ID, dbg_q);
>> + return 0;
>> + }
>> +
>> + q_table->kernel_vaddr = dma_alloc_attrs(dev, ALIGNED_QUEUE_SIZE,
>> + &q_table->device_addr,
>> + GFP_KERNEL, DMA_ATTR_WRITE_COMBINE);
>> + if (!q_table->kernel_vaddr) {
>> + dev_err(dev, "%s: queues alloc and map failed\n", __func__);
>> + return -ENOMEM;
>> + }
>> +
>> + sfr->kernel_vaddr = dma_alloc_attrs(dev, ALIGNED_SFR_SIZE,
>> + &sfr->device_addr,
>> + GFP_KERNEL, DMA_ATTR_WRITE_COMBINE);
>> + if (!sfr->kernel_vaddr) {
>> + dev_err(dev, "%s: sfr alloc and map failed\n", __func__);
>> + return -ENOMEM;
>> + }
>> +
>> + q_table->size = IFACEQ_TABLE_SIZE;
>> +
>> + __queue_init(q_table, IFACEQ_CMDQ_ID, cmd_q);
>> + __queue_init(q_table, IFACEQ_MSGQ_ID, msg_q);
>> + __queue_init(q_table, IFACEQ_DBGQ_ID, dbg_q);
>> +
>> + q_tbl_hdr = (struct hfi_queue_table_header *)q_table->kernel_vaddr;
>> + q_tbl_hdr->qtbl_version = 0;
>> + q_tbl_hdr->device_addr = (void *)daddr;
>> + strscpy(q_tbl_hdr->name, "hfi-queues", sizeof(q_tbl_hdr->name));
>> + q_tbl_hdr->qtbl_size = IFACEQ_TABLE_SIZE;
>> + q_tbl_hdr->qtbl_qhdr0_offset = sizeof(*q_tbl_hdr);
>> + q_tbl_hdr->qtbl_qhdr_size = sizeof(struct hfi_queue_header);
>> + q_tbl_hdr->qtbl_num_q = IFACEQ_NUMQ;
>> + q_tbl_hdr->qtbl_num_active_q = IFACEQ_NUMQ;
>> +
>> + sfr->size = ALIGNED_SFR_SIZE;
>> + /* Write sfr size in first word to be used by firmware */
>> + *((u32 *)sfr->kernel_vaddr) = sfr->size;
>> +
>> + return 0;
>> +}
>> +
>> +static void __queue_deinit(struct iface_q_info *iface_q)
>> +{
>> + iface_q->qhdr = NULL;
>> + iface_q->q_array.kernel_vaddr = NULL;
>> + iface_q->q_array.device_addr = 0;
>> +}
>> +
>> +void hfi_queue_deinit(struct device *dev, struct mem_desc *q_table,
>> struct mem_desc *sfr,
>> + struct iface_q_info *cmd_q, struct iface_q_info *msg_q,
>> + struct iface_q_info *dbg_q)
>> +{
>> + if (!q_table->kernel_vaddr)
>> + return;
>> +
>> + dma_free_attrs(dev, q_table->size, q_table->kernel_vaddr,
>> + q_table->device_addr, q_table->attrs);
>> +
>> + dma_free_attrs(dev, sfr->size, sfr->kernel_vaddr,
>> + sfr->device_addr, sfr->attrs);
>> +
>> + __queue_deinit(cmd_q);
>> + __queue_deinit(msg_q);
>> + __queue_deinit(dbg_q);
>> +
>> + q_table->kernel_vaddr = NULL;
>> + q_table->device_addr = 0;
>> +
>> + sfr->kernel_vaddr = NULL;
>> + sfr->device_addr = 0;
>> +}
>> diff --git a/drivers/media/platform/qcom/vcodec/hfi_queue.h
>> b/drivers/media/platform/qcom/vcodec/hfi_queue.h
>> new file mode 100644
>> index 0000000..8cc9491
>> --- /dev/null
>> +++ b/drivers/media/platform/qcom/vcodec/hfi_queue.h
>> @@ -0,0 +1,129 @@
>> +/* SPDX-License-Identifier: GPL-2.0-only */
>> +/*
>> + * Copyright (c) 2022-2023 Qualcomm Innovation Center, Inc. All rights
>> reserved.
>> + */
>> +
>> +#ifndef _HFI_QUEUE_H_
>> +#define _HFI_QUEUE_H_
>> +
>> +#include <linux/types.h>
>> +#include <media/v4l2-device.h>
>> +
>> +/*
>> + * Shared queues are used for communication between driver and firmware.
>> + * There are 3 types of queues:
>> + * Command queue - driver to write any command to firmware.
>> + * Message queue - firmware to send any response to driver.
>> + * Debug queue - firmware to write debug message.
>> + */
>> +
>> +/* Host-firmware shared queue ids */
>> +enum iris_iface_queue {
>> + IFACEQ_CMDQ_ID = 0,
>> + IFACEQ_MSGQ_ID,
>> + IFACEQ_DBGQ_ID,
>> + IFACEQ_NUMQ, /* not an index */
>> +};
>> +
>> +#define IFACEQ_MAX_BUF_COUNT 50
>> +/*
>> + * Max session supported are 16.
>> + * this value is used to calculate the size of
>> + * individual shared queue.
>> + */
>> +#define IFACE_MAX_PARALLEL_CLNTS 16
>> +#define IFACEQ_DFLT_QHDR 0x01010000
>> +
>> +#define IFACEQ_MAX_PKT_SIZE 1024
>> +#define IFACEQ_CORE_PKT_SIZE (1024 * 4)
>> +
>> +#define IFACEQ_TABLE_SIZE (sizeof(struct hfi_queue_table_header) + \
>> + sizeof(struct hfi_queue_header) * IFACEQ_NUMQ)
>> +
>> +#define IFACEQ_QUEUE_SIZE (IFACEQ_MAX_PKT_SIZE * \
>> + IFACEQ_MAX_BUF_COUNT * IFACE_MAX_PARALLEL_CLNTS)
>> +
>> +#define IFACEQ_GET_QHDR_START_ADDR(ptr, i) \
>> + ((void *)(((ptr) + sizeof(struct hfi_queue_table_header)) + \
>> + ((i) * sizeof(struct hfi_queue_header))))
>> +
>> +#define SFR_SIZE SZ_4K
>> +#define QUEUE_SIZE (IFACEQ_TABLE_SIZE + \
>> + (IFACEQ_QUEUE_SIZE * IFACEQ_NUMQ))
>> +
>> +#define ALIGNED_SFR_SIZE ALIGN(SFR_SIZE, SZ_4K)
>> +#define ALIGNED_QUEUE_SIZE ALIGN(QUEUE_SIZE, SZ_4K)
>> +#define SHARED_QSIZE ALIGN(ALIGNED_SFR_SIZE + ALIGNED_QUEUE_SIZE,
>> SZ_1M)
>> +
>> +struct mem_desc {
>> + dma_addr_t device_addr;
>> + void *kernel_vaddr;
>> + u32 size;
>> + unsigned long attrs;
>> +};
>> +
>> +struct hfi_queue_table_header {
>> + u32 qtbl_version;
>> + u32 qtbl_size;
>> + u32 qtbl_qhdr0_offset;
>> + u32 qtbl_qhdr_size;
>> + u32 qtbl_num_q;
>> + u32 qtbl_num_active_q;
>> + void *device_addr;
>> + char name[256];
>> +};
>> +
>> +/**
>> + * struct hfi_queue_header - queue header definition
>> + *
>> + * @qhdr_status: bits (7:0) 0x1 - active 0x0 - inactive
>> + * @qhdr_start_addr: queue start addr.
>> + * @qhdr_type: queue type.
>> + * @qhdr_q_size: size in number of HFI_QUEUE_PACKET_TYPE packets it can
>> hold.
>> + * @qhdr_pkt_size: size of the Queue packet entries, in bytes.
>> + * @qhdr_rx_wm: the receiver watermark in number of queue packets.
>> + * @qhdr_tx_wm: the transmitter watermark in number of queue packets.
>> + * @qhdr_rx_req: set to request an interrupt from transmitter if
>> qhdr_tx_wm is reached.
>> + * @qhdr_tx_req: set to request an interrupt from receiver if qhdr_rx_wm
>> is reached.
>> + * @qhdr_rx_irq_status: status of receiver irq.
>> + * @qhdr_tx_irq_status: status of transmitter irq.
>> + * @qhdr_read_idx: read index, which starts with a value of 0 and wraps
>> around after
>> + * (QHDR_Q_SIZE - 1).
>> + * @qhdr_write_idx: write index, which starts with a value of 0 and
>> wraps around after
>> + * (QHDR_Q_SIZE - 1).
>> + */
>> +
>> +struct hfi_queue_header {
>> + u32 qhdr_status;
>> + u32 qhdr_start_addr;
>> + u32 qhdr_type;
>> + u32 qhdr_q_size;
>> + u32 qhdr_pkt_size;
>> + u32 qhdr_pkt_drop_cnt;
>> + u32 qhdr_rx_wm;
>> + u32 qhdr_tx_wm;
>> + u32 qhdr_rx_req;
>> + u32 qhdr_tx_req;
>> + u32 qhdr_rx_irq_status;
>> + u32 qhdr_tx_irq_status;
>> + u32 qhdr_read_idx;
>> + u32 qhdr_write_idx;
>> +};
>> +
>> +struct iface_q_info {
>> + struct hfi_queue_header *qhdr;
>> + struct mem_desc q_array;
>> +};
>> +
>> +int write_queue(struct iface_q_info *qinfo, void *packet,
>> + u32 packet_size, u32 *rx_req);
>> +int read_queue(struct iface_q_info *qinfo, void *packet, u32 *tx_req);
>> +int hfi_queue_init(struct device *dev, struct mem_desc *q_table, struct
>> mem_desc *sfr,
>> + struct iface_q_info *cmd_q, struct iface_q_info *msg_q,
>> + struct iface_q_info *dbg_q, void *daddr);
>> +
>> +void hfi_queue_deinit(struct device *dev, struct mem_desc *q_table,
>> struct mem_desc *sfr,
>> + struct iface_q_info *cmd_q, struct iface_q_info *msg_q,
>> + struct iface_q_info *dbg_q);
>> +
>> +#endif
>> diff --git a/drivers/media/platform/qcom/vcodec/venus/Makefile
>> b/drivers/media/platform/qcom/vcodec/venus/Makefile
>> index f6f3a88..1941ef4 100644
>> --- a/drivers/media/platform/qcom/vcodec/venus/Makefile
>> +++ b/drivers/media/platform/qcom/vcodec/venus/Makefile
>> @@ -1,7 +1,8 @@
>> # SPDX-License-Identifier: GPL-2.0
>> # Makefile for Qualcomm Venus driver
>> -venus-core-objs += ../firmware.o
>> +venus-core-objs += ../firmware.o \
>> + ../hfi_queue.o
>> venus-core-objs += core.o helpers.o firmware_no_tz.o \
>> hfi_venus.o hfi_msgs.o hfi_cmds.o hfi.o \
>> diff --git a/drivers/media/platform/qcom/vcodec/venus/hfi_venus.c
>> b/drivers/media/platform/qcom/vcodec/venus/hfi_venus.c
>> index 5a68db9..41e3f27 100644
>> --- a/drivers/media/platform/qcom/vcodec/venus/hfi_venus.c
>> +++ b/drivers/media/platform/qcom/vcodec/venus/hfi_venus.c
>> @@ -16,6 +16,7 @@
>> #include "../firmware.h"
>> #include "hfi_cmds.h"
>> #include "hfi_msgs.h"
>> +#include "../hfi_queue.h"
>> #include "hfi_venus.h"
>> #include "hfi_venus_io.h"
>> #include "firmware_no_tz.h"
>> @@ -34,79 +35,13 @@
>> #define IFACEQ_CMD_IDX 0
>> #define IFACEQ_MSG_IDX 1
>> #define IFACEQ_DBG_IDX 2
>> -#define IFACEQ_MAX_BUF_COUNT 50
>> -#define IFACEQ_MAX_PARALLEL_CLNTS 16
>> -#define IFACEQ_DFLT_QHDR 0x01010000
>> #define POLL_INTERVAL_US 50
>> -#define IFACEQ_MAX_PKT_SIZE 1024
>> -#define IFACEQ_MED_PKT_SIZE 768
>> -#define IFACEQ_MIN_PKT_SIZE 8
>> #define IFACEQ_VAR_SMALL_PKT_SIZE 100
>> #define IFACEQ_VAR_LARGE_PKT_SIZE 512
>> #define IFACEQ_VAR_HUGE_PKT_SIZE (1024 * 12)
>> -struct hfi_queue_table_header {
>> - u32 version;
>> - u32 size;
>> - u32 qhdr0_offset;
>> - u32 qhdr_size;
>> - u32 num_q;
>> - u32 num_active_q;
>> -};
>> -
>> -struct hfi_queue_header {
>> - u32 status;
>> - u32 start_addr;
>> - u32 type;
>> - u32 q_size;
>> - u32 pkt_size;
>> - u32 pkt_drop_cnt;
>> - u32 rx_wm;
>> - u32 tx_wm;
>> - u32 rx_req;
>> - u32 tx_req;
>> - u32 rx_irq_status;
>> - u32 tx_irq_status;
>> - u32 read_idx;
>> - u32 write_idx;
>> -};
>> -
>> -#define IFACEQ_TABLE_SIZE \
>> - (sizeof(struct hfi_queue_table_header) + \
>> - sizeof(struct hfi_queue_header) * IFACEQ_NUM)
>> -
>> -#define IFACEQ_QUEUE_SIZE (IFACEQ_MAX_PKT_SIZE * \
>> - IFACEQ_MAX_BUF_COUNT * IFACEQ_MAX_PARALLEL_CLNTS)
>> -
>> -#define IFACEQ_GET_QHDR_START_ADDR(ptr, i) \
>> - (void *)(((ptr) + sizeof(struct hfi_queue_table_header)) + \
>> - ((i) * sizeof(struct hfi_queue_header)))
>> -
>> -#define QDSS_SIZE SZ_4K
>> -#define SFR_SIZE SZ_4K
>> -#define QUEUE_SIZE \
>> - (IFACEQ_TABLE_SIZE + (IFACEQ_QUEUE_SIZE * IFACEQ_NUM))
>> -
>> -#define ALIGNED_QDSS_SIZE ALIGN(QDSS_SIZE, SZ_4K)
>> -#define ALIGNED_SFR_SIZE ALIGN(SFR_SIZE, SZ_4K)
>> -#define ALIGNED_QUEUE_SIZE ALIGN(QUEUE_SIZE, SZ_4K)
>> -#define SHARED_QSIZE ALIGN(ALIGNED_SFR_SIZE + ALIGNED_QUEUE_SIZE + \
>> - ALIGNED_QDSS_SIZE, SZ_1M)
>> -
>> -struct mem_desc {
>> - dma_addr_t da; /* device address */
>> - void *kva; /* kernel virtual address */
>> - u32 size;
>> - unsigned long attrs;
>> -};
>> -
>> -struct iface_queue {
>> - struct hfi_queue_header *qhdr;
>> - struct mem_desc qmem;
>> -};
>> -
>> enum venus_state {
>> VENUS_STATE_DEINIT = 1,
>> VENUS_STATE_INIT,
>> @@ -125,12 +60,11 @@ struct venus_hfi_device {
>> struct completion release_resource;
>> struct mem_desc ifaceq_table;
>> struct mem_desc sfr;
>> - struct iface_queue queues[IFACEQ_NUM];
>> + struct iface_q_info queues[IFACEQ_NUM];
>> u8 pkt_buf[IFACEQ_VAR_HUGE_PKT_SIZE];
>> u8 dbg_buf[IFACEQ_VAR_HUGE_PKT_SIZE];
>> };
>> -static bool venus_pkt_debug;
>> int venus_fw_debug = HFI_DEBUG_MSG_ERROR | HFI_DEBUG_MSG_FATAL;
>> static bool venus_fw_low_power_mode = true;
>> static int venus_hw_rsp_timeout = 1000;
>> @@ -149,212 +83,6 @@ static bool venus_is_valid_state(struct
>> venus_hfi_device *hdev)
>> return hdev->state != VENUS_STATE_DEINIT;
>> }
>> -static void venus_dump_packet(struct venus_hfi_device *hdev, const
>> void *packet)
>> -{
>> - size_t pkt_size = *(u32 *)packet;
>> -
>> - if (!venus_pkt_debug)
>> - return;
>> -
>> - print_hex_dump(KERN_DEBUG, "", DUMP_PREFIX_OFFSET, 16, 1, packet,
>> - pkt_size, true);
>> -}
>> -
>> -static int venus_write_queue(struct venus_hfi_device *hdev,
>> - struct iface_queue *queue,
>> - void *packet, u32 *rx_req)
>> -{
>> - struct hfi_queue_header *qhdr;
>> - u32 dwords, new_wr_idx;
>> - u32 empty_space, rd_idx, wr_idx, qsize;
>> - u32 *wr_ptr;
>> -
>> - if (!queue->qmem.kva)
>> - return -EINVAL;
>> -
>> - qhdr = queue->qhdr;
>> - if (!qhdr)
>> - return -EINVAL;
>> -
>> - venus_dump_packet(hdev, packet);
>> -
>> - dwords = (*(u32 *)packet) >> 2;
>> - if (!dwords)
>> - return -EINVAL;
>> -
>> - rd_idx = qhdr->read_idx;
>> - wr_idx = qhdr->write_idx;
>> - qsize = qhdr->q_size;
>> - /* ensure rd/wr indices's are read from memory */
>> - rmb();
>> -
>> - if (wr_idx >= rd_idx)
>> - empty_space = qsize - (wr_idx - rd_idx);
>> - else
>> - empty_space = rd_idx - wr_idx;
>> -
>> - if (empty_space <= dwords) {
>> - qhdr->tx_req = 1;
>> - /* ensure tx_req is updated in memory */
>> - wmb();
>> - return -ENOSPC;
>> - }
>> -
>> - qhdr->tx_req = 0;
>> - /* ensure tx_req is updated in memory */
>> - wmb();
>> -
>> - new_wr_idx = wr_idx + dwords;
>> - wr_ptr = (u32 *)(queue->qmem.kva + (wr_idx << 2));
>> -
>> - if (wr_ptr < (u32 *)queue->qmem.kva ||
>> - wr_ptr > (u32 *)(queue->qmem.kva + queue->qmem.size -
>> sizeof(*wr_ptr)))
>> - return -EINVAL;
>> -
>> - if (new_wr_idx < qsize) {
>> - memcpy(wr_ptr, packet, dwords << 2);
>> - } else {
>> - size_t len;
>> -
>> - new_wr_idx -= qsize;
>> - len = (dwords - new_wr_idx) << 2;
>> - memcpy(wr_ptr, packet, len);
>> - memcpy(queue->qmem.kva, packet + len, new_wr_idx << 2);
>> - }
>> -
>> - /* make sure packet is written before updating the write index */
>> - wmb();
>> -
>> - qhdr->write_idx = new_wr_idx;
>> - *rx_req = qhdr->rx_req ? 1 : 0;
>> -
>> - /* make sure write index is updated before an interrupt is raised */
>> - mb();
>> -
>> - return 0;
>> -}
>> -
>> -static int venus_read_queue(struct venus_hfi_device *hdev,
>> - struct iface_queue *queue, void *pkt, u32 *tx_req)
>> -{
>> - struct hfi_queue_header *qhdr;
>> - u32 dwords, new_rd_idx;
>> - u32 rd_idx, wr_idx, type, qsize;
>> - u32 *rd_ptr;
>> - u32 recv_request = 0;
>> - int ret = 0;
>> -
>> - if (!queue->qmem.kva)
>> - return -EINVAL;
>> -
>> - qhdr = queue->qhdr;
>> - if (!qhdr)
>> - return -EINVAL;
>> -
>> - type = qhdr->type;
>> - rd_idx = qhdr->read_idx;
>> - wr_idx = qhdr->write_idx;
>> - qsize = qhdr->q_size;
>> -
>> - /* make sure data is valid before using it */
>> - rmb();
>> -
>> - /*
>> - * Do not set receive request for debug queue, if set, Venus generates
>> - * interrupt for debug messages even when there is no response message
>> - * available. In general debug queue will not become full as it is
>> being
>> - * emptied out for every interrupt from Venus. Venus will anyway
>> - * generates interrupt if it is full.
>> - */
>> - if (type & HFI_CTRL_TO_HOST_MSG_Q)
>> - recv_request = 1;
>> -
>> - if (rd_idx == wr_idx) {
>> - qhdr->rx_req = recv_request;
>> - *tx_req = 0;
>> - /* update rx_req field in memory */
>> - wmb();
>> - return -ENODATA;
>> - }
>> -
>> - rd_ptr = (u32 *)(queue->qmem.kva + (rd_idx << 2));
>> -
>> - if (rd_ptr < (u32 *)queue->qmem.kva ||
>> - rd_ptr > (u32 *)(queue->qmem.kva + queue->qmem.size -
>> sizeof(*rd_ptr)))
>> - return -EINVAL;
>> -
>> - dwords = *rd_ptr >> 2;
>> - if (!dwords)
>> - return -EINVAL;
>> -
>> - new_rd_idx = rd_idx + dwords;
>> - if (((dwords << 2) <= IFACEQ_VAR_HUGE_PKT_SIZE) && rd_idx <= qsize) {
>> - if (new_rd_idx < qsize) {
>> - memcpy(pkt, rd_ptr, dwords << 2);
>> - } else {
>> - size_t len;
>> -
>> - new_rd_idx -= qsize;
>> - len = (dwords - new_rd_idx) << 2;
>> - memcpy(pkt, rd_ptr, len);
>> - memcpy(pkt + len, queue->qmem.kva, new_rd_idx << 2);
>> - }
>> - } else {
>> - /* bad packet received, dropping */
>> - new_rd_idx = qhdr->write_idx;
>> - ret = -EBADMSG;
>> - }
>> -
>> - /* ensure the packet is read before updating read index */
>> - rmb();
>> -
>> - qhdr->read_idx = new_rd_idx;
>> - /* ensure updating read index */
>> - wmb();
>> -
>> - rd_idx = qhdr->read_idx;
>> - wr_idx = qhdr->write_idx;
>> - /* ensure rd/wr indices are read from memory */
>> - rmb();
>> -
>> - if (rd_idx != wr_idx)
>> - qhdr->rx_req = 0;
>> - else
>> - qhdr->rx_req = recv_request;
>> -
>> - *tx_req = qhdr->tx_req ? 1 : 0;
>> -
>> - /* ensure rx_req is stored to memory and tx_req is loaded from
>> memory */
>> - mb();
>> -
>> - venus_dump_packet(hdev, pkt);
>> -
>> - return ret;
>> -}
>> -
>> -static int venus_alloc(struct venus_hfi_device *hdev, struct mem_desc
>> *desc,
>> - u32 size)
>> -{
>> - struct device *dev = hdev->core->dev;
>> -
>> - desc->attrs = DMA_ATTR_WRITE_COMBINE;
>> - desc->size = ALIGN(size, SZ_4K);
>> -
>> - desc->kva = dma_alloc_attrs(dev, desc->size, &desc->da, GFP_KERNEL,
>> - desc->attrs);
>> - if (!desc->kva)
>> - return -ENOMEM;
>> -
>> - return 0;
>> -}
>> -
>> -static void venus_free(struct venus_hfi_device *hdev, struct mem_desc *mem)
>> -{
>> - struct device *dev = hdev->core->dev;
>> -
>> - dma_free_attrs(dev, mem->size, mem->kva, mem->da, mem->attrs);
>> -}
>> -
>> static void venus_set_registers(struct venus_hfi_device *hdev)
>> {
>> const struct venus_resources *res = hdev->core->res;
>> @@ -380,11 +108,11 @@ static void venus_soft_int(struct venus_hfi_device
>> *hdev)
>> }
>> static int venus_iface_cmdq_write_nolock(struct venus_hfi_device *hdev,
>> - void *pkt, bool sync)
>> + void *pkt, u32 pkt_size, bool sync)
>
> I'm tempted to say that you should first move the code, then add pkt_size
> later. BTW, why do we need it at all and why wasn't it present beforehand?
>
This was added to make the code generic between venus and iris.
But will check if I can move the code first and then add additional
parameters with iris changes.
>> {
>> struct device *dev = hdev->core->dev;
>> struct hfi_pkt_hdr *cmd_packet;
>> - struct iface_queue *queue;
>> + struct iface_q_info *queue;
>
> iface_queue was good enough, unless you have a reason.
>
Sure, will fix.
>> u32 rx_req;
>> int ret;
>> @@ -396,7 +124,7 @@ static int venus_iface_cmdq_write_nolock(struct
>> venus_hfi_device *hdev,
>> queue = &hdev->queues[IFACEQ_CMD_IDX];
>> - ret = venus_write_queue(hdev, queue, pkt, &rx_req);
>> + ret = write_queue(queue, pkt, pkt_size, &rx_req);
>
> You know, you can call venus_write_queue from your Iris driver, the kernel
> won't bite you. But the generic write_queue() function will be bitten and
> frowned upon.
>
We would prefer to keep the common APIs like this in common vcodec folder,
with proper prefix for sure and use that in both venus and iris driver.
>> if (ret) {
>> dev_err(dev, "write to iface cmd queue failed (%d)\n", ret);
>> return ret;
>> @@ -408,7 +136,7 @@ static int venus_iface_cmdq_write_nolock(struct
>> venus_hfi_device *hdev,
>> * commands
>> */
>> queue = &hdev->queues[IFACEQ_MSG_IDX];
>> - queue->qhdr->rx_req = 1;
>> + queue->qhdr->qhdr_rx_req = 1;
>> /* ensure rx_req is updated in memory */
>> wmb();
>> }
>> @@ -419,12 +147,12 @@ static int venus_iface_cmdq_write_nolock(struct
>> venus_hfi_device *hdev,
>> return 0;
>> }
>> -static int venus_iface_cmdq_write(struct venus_hfi_device *hdev, void
>> *pkt, bool sync)
>> +static int venus_iface_cmdq_write(struct venus_hfi_device *hdev, void
>> *pkt, u32 pkt_size, bool sync)
>> {
>> int ret;
>> mutex_lock(&hdev->lock);
>> - ret = venus_iface_cmdq_write_nolock(hdev, pkt, sync);
>> + ret = venus_iface_cmdq_write_nolock(hdev, pkt, pkt_size, sync);
>> mutex_unlock(&hdev->lock);
>> return ret;
>> @@ -447,7 +175,7 @@ static int venus_hfi_core_set_resource(struct
>> venus_core *core, u32 id,
>> if (ret)
>> return ret;
>> - ret = venus_iface_cmdq_write(hdev, pkt, false);
>> + ret = venus_iface_cmdq_write(hdev, pkt, pkt->hdr.size, false);
>> if (ret)
>> return ret;
>> @@ -531,12 +259,12 @@ static int venus_run(struct venus_hfi_device *hdev)
>> */
>> venus_set_registers(hdev);
>> - writel(hdev->ifaceq_table.da, cpu_cs_base + UC_REGION_ADDR);
>> + writel(hdev->ifaceq_table.device_addr, cpu_cs_base + UC_REGION_ADDR);
>> writel(SHARED_QSIZE, cpu_cs_base + UC_REGION_SIZE);
>> - writel(hdev->ifaceq_table.da, cpu_cs_base + CPU_CS_SCIACMDARG2);
>> + writel(hdev->ifaceq_table.device_addr, cpu_cs_base +
>> CPU_CS_SCIACMDARG2);
>
> This is a separate patch material.
>
Sure, will try to separate these changes.
>> writel(0x01, cpu_cs_base + CPU_CS_SCIACMDARG1);
>> - if (hdev->sfr.da)
>> - writel(hdev->sfr.da, cpu_cs_base + SFR_ADDR);
>> + if (hdev->sfr.device_addr)
>> + writel(hdev->sfr.device_addr, cpu_cs_base + SFR_ADDR);
>> ret = venus_boot_core(hdev);
>> if (ret) {
>> @@ -690,7 +418,7 @@ static int venus_power_on(struct venus_hfi_device *hdev)
>> static int venus_iface_msgq_read_nolock(struct venus_hfi_device *hdev,
>> void *pkt)
>> {
>> - struct iface_queue *queue;
>> + struct iface_q_info *queue;
>> u32 tx_req;
>> int ret;
>> @@ -699,7 +427,7 @@ static int venus_iface_msgq_read_nolock(struct
>> venus_hfi_device *hdev,
>> queue = &hdev->queues[IFACEQ_MSG_IDX];
>> - ret = venus_read_queue(hdev, queue, pkt, &tx_req);
>> + ret = read_queue(queue, pkt, &tx_req);
>> if (ret)
>> return ret;
>> @@ -723,7 +451,7 @@ static int venus_iface_msgq_read(struct
>> venus_hfi_device *hdev, void *pkt)
>> static int venus_iface_dbgq_read_nolock(struct venus_hfi_device *hdev,
>> void *pkt)
>> {
>> - struct iface_queue *queue;
>> + struct iface_q_info *queue;
>> u32 tx_req;
>> int ret;
>> @@ -733,7 +461,7 @@ static int venus_iface_dbgq_read_nolock(struct
>> venus_hfi_device *hdev,
>> queue = &hdev->queues[IFACEQ_DBG_IDX];
>> - ret = venus_read_queue(hdev, queue, pkt, &tx_req);
>> + ret = read_queue(queue, pkt, &tx_req);
>> if (ret)
>> return ret;
>> @@ -759,100 +487,18 @@ static int venus_iface_dbgq_read(struct
>> venus_hfi_device *hdev, void *pkt)
>> static void venus_set_qhdr_defaults(struct hfi_queue_header *qhdr)
>> {
>> - qhdr->status = 1;
>> - qhdr->type = IFACEQ_DFLT_QHDR;
>> - qhdr->q_size = IFACEQ_QUEUE_SIZE / 4;
>> - qhdr->pkt_size = 0;
>> - qhdr->rx_wm = 1;
>> - qhdr->tx_wm = 1;
>> - qhdr->rx_req = 1;
>> - qhdr->tx_req = 0;
>> - qhdr->rx_irq_status = 0;
>> - qhdr->tx_irq_status = 0;
>> - qhdr->read_idx = 0;
>> - qhdr->write_idx = 0;
>> -}
>> -
>> -static void venus_interface_queues_release(struct venus_hfi_device *hdev)
>> -{
>> - mutex_lock(&hdev->lock);
>> -
>> - venus_free(hdev, &hdev->ifaceq_table);
>> - venus_free(hdev, &hdev->sfr);
>> -
>> - memset(hdev->queues, 0, sizeof(hdev->queues));
>> - memset(&hdev->ifaceq_table, 0, sizeof(hdev->ifaceq_table));
>> - memset(&hdev->sfr, 0, sizeof(hdev->sfr));
>> -
>> - mutex_unlock(&hdev->lock);
>> -}
>> -
>> -static int venus_interface_queues_init(struct venus_hfi_device *hdev)
>> -{
>> - struct hfi_queue_table_header *tbl_hdr;
>> - struct iface_queue *queue;
>> - struct hfi_sfr *sfr;
>> - struct mem_desc desc = {0};
>> - unsigned int offset;
>> - unsigned int i;
>> - int ret;
>> -
>> - ret = venus_alloc(hdev, &desc, ALIGNED_QUEUE_SIZE);
>> - if (ret)
>> - return ret;
>> -
>> - hdev->ifaceq_table = desc;
>> - offset = IFACEQ_TABLE_SIZE;
>> -
>> - for (i = 0; i < IFACEQ_NUM; i++) {
>> - queue = &hdev->queues[i];
>> - queue->qmem.da = desc.da + offset;
>> - queue->qmem.kva = desc.kva + offset;
>> - queue->qmem.size = IFACEQ_QUEUE_SIZE;
>> - offset += queue->qmem.size;
>> - queue->qhdr =
>> - IFACEQ_GET_QHDR_START_ADDR(hdev->ifaceq_table.kva, i);
>> -
>> - venus_set_qhdr_defaults(queue->qhdr);
>> -
>> - queue->qhdr->start_addr = queue->qmem.da;
>> -
>> - if (i == IFACEQ_CMD_IDX)
>> - queue->qhdr->type |= HFI_HOST_TO_CTRL_CMD_Q;
>> - else if (i == IFACEQ_MSG_IDX)
>> - queue->qhdr->type |= HFI_CTRL_TO_HOST_MSG_Q;
>> - else if (i == IFACEQ_DBG_IDX)
>> - queue->qhdr->type |= HFI_CTRL_TO_HOST_DBG_Q;
>> - }
>> -
>> - tbl_hdr = hdev->ifaceq_table.kva;
>> - tbl_hdr->version = 0;
>> - tbl_hdr->size = IFACEQ_TABLE_SIZE;
>> - tbl_hdr->qhdr0_offset = sizeof(struct hfi_queue_table_header);
>> - tbl_hdr->qhdr_size = sizeof(struct hfi_queue_header);
>> - tbl_hdr->num_q = IFACEQ_NUM;
>> - tbl_hdr->num_active_q = IFACEQ_NUM;
>> -
>> - /*
>> - * Set receive request to zero on debug queue as there is no
>> - * need of interrupt from video hardware for debug messages
>> - */
>> - queue = &hdev->queues[IFACEQ_DBG_IDX];
>> - queue->qhdr->rx_req = 0;
>> -
>> - ret = venus_alloc(hdev, &desc, ALIGNED_SFR_SIZE);
>> - if (ret) {
>> - hdev->sfr.da = 0;
>> - } else {
>> - hdev->sfr = desc;
>> - sfr = hdev->sfr.kva;
>> - sfr->buf_size = ALIGNED_SFR_SIZE;
>> - }
>> -
>> - /* ensure table and queue header structs are settled in memory */
>> - wmb();
>> -
>> - return 0;
>> + qhdr->qhdr_status = 1;
>> + qhdr->qhdr_type = IFACEQ_DFLT_QHDR;
>> + qhdr->qhdr_q_size = IFACEQ_QUEUE_SIZE / 4;
>> + qhdr->qhdr_pkt_size = 0;
>> + qhdr->qhdr_rx_wm = 1;
>> + qhdr->qhdr_tx_wm = 1;
>> + qhdr->qhdr_rx_req = 1;
>> + qhdr->qhdr_tx_req = 0;
>> + qhdr->qhdr_rx_irq_status = 0;
>> + qhdr->qhdr_tx_irq_status = 0;
>> + qhdr->qhdr_read_idx = 0;
>> + qhdr->qhdr_write_idx = 0;
>> }
>> static int venus_sys_set_debug(struct venus_hfi_device *hdev, u32 debug)
>> @@ -864,7 +510,7 @@ static int venus_sys_set_debug(struct
>> venus_hfi_device *hdev, u32 debug)
>> pkt_sys_debug_config(pkt, HFI_DEBUG_MODE_QUEUE, debug);
>> - return venus_iface_cmdq_write(hdev, pkt, false);
>> + return venus_iface_cmdq_write(hdev, pkt, pkt->hdr.size, false);
>> }
>> static int venus_sys_set_coverage(struct venus_hfi_device *hdev, u32
>> mode)
>> @@ -876,7 +522,7 @@ static int venus_sys_set_coverage(struct
>> venus_hfi_device *hdev, u32 mode)
>> pkt_sys_coverage_config(pkt, mode);
>> - return venus_iface_cmdq_write(hdev, pkt, false);
>> + return venus_iface_cmdq_write(hdev, pkt, pkt->hdr.size, false);
>> }
>> static int venus_sys_set_idle_message(struct venus_hfi_device *hdev,
>> @@ -892,7 +538,7 @@ static int venus_sys_set_idle_message(struct
>> venus_hfi_device *hdev,
>> pkt_sys_idle_indicator(pkt, enable);
>> - return venus_iface_cmdq_write(hdev, pkt, false);
>> + return venus_iface_cmdq_write(hdev, pkt, pkt->hdr.size, false);
>> }
>> static int venus_sys_set_power_control(struct venus_hfi_device *hdev,
>> @@ -905,7 +551,7 @@ static int venus_sys_set_power_control(struct
>> venus_hfi_device *hdev,
>> pkt_sys_power_control(pkt, enable);
>> - return venus_iface_cmdq_write(hdev, pkt, false);
>> + return venus_iface_cmdq_write(hdev, pkt, pkt->hdr.size, false);
>> }
>> static int venus_sys_set_ubwc_config(struct venus_hfi_device *hdev)
>> @@ -919,7 +565,7 @@ static int venus_sys_set_ubwc_config(struct
>> venus_hfi_device *hdev)
>> pkt_sys_ubwc_config(pkt, res->ubwc_conf);
>> - ret = venus_iface_cmdq_write(hdev, pkt, false);
>> + ret = venus_iface_cmdq_write(hdev, pkt, pkt->hdr.size, false);
>> if (ret)
>> return ret;
>> @@ -938,7 +584,7 @@ static int venus_get_queue_size(struct
>> venus_hfi_device *hdev,
>> if (!qhdr)
>> return -EINVAL;
>> - return abs(qhdr->read_idx - qhdr->write_idx);
>> + return abs(qhdr->qhdr_read_idx - qhdr->qhdr_write_idx);
>> }
>> static int venus_sys_set_default_properties(struct venus_hfi_device
>> *hdev)
>> @@ -980,7 +626,7 @@ static int venus_session_cmd(struct venus_inst *inst,
>> u32 pkt_type, bool sync)
>> pkt_session_cmd(&pkt, pkt_type, inst);
>> - return venus_iface_cmdq_write(hdev, &pkt, sync);
>> + return venus_iface_cmdq_write(hdev, &pkt, pkt.shdr.hdr.size, sync);
>> }
>> static void venus_flush_debug_queue(struct venus_hfi_device *hdev)
>> @@ -1010,7 +656,7 @@ static int venus_prepare_power_collapse(struct
>> venus_hfi_device *hdev,
>> pkt_sys_pc_prep(&pkt);
>> - ret = venus_iface_cmdq_write(hdev, &pkt, false);
>> + ret = venus_iface_cmdq_write(hdev, &pkt, pkt.hdr.size, false);
>> if (ret)
>> return ret;
>> @@ -1047,7 +693,7 @@ static int venus_are_queues_empty(struct
>> venus_hfi_device *hdev)
>> static void venus_sfr_print(struct venus_hfi_device *hdev)
>> {
>> struct device *dev = hdev->core->dev;
>> - struct hfi_sfr *sfr = hdev->sfr.kva;
>> + struct hfi_sfr *sfr = hdev->sfr.kernel_vaddr;
>> void *p;
>> if (!sfr)
>> @@ -1163,13 +809,13 @@ static int venus_core_init(struct venus_core *core)
>> venus_set_state(hdev, VENUS_STATE_INIT);
>> - ret = venus_iface_cmdq_write(hdev, &pkt, false);
>> + ret = venus_iface_cmdq_write(hdev, &pkt, pkt.hdr.size, false);
>> if (ret)
>> return ret;
>> pkt_sys_image_version(&version_pkt);
>> - ret = venus_iface_cmdq_write(hdev, &version_pkt, false);
>> + ret = venus_iface_cmdq_write(hdev, &version_pkt,
>> version_pkt.hdr.size, false);
>> if (ret)
>> dev_warn(dev, "failed to send image version pkt to fw\n");
>> @@ -1198,7 +844,7 @@ static int venus_core_ping(struct venus_core
>> *core, u32 cookie)
>> pkt_sys_ping(&pkt, cookie);
>> - return venus_iface_cmdq_write(hdev, &pkt, false);
>> + return venus_iface_cmdq_write(hdev, &pkt, pkt.hdr.size, false);
>> }
>> static int venus_core_trigger_ssr(struct venus_core *core, u32
>> trigger_type)
>> @@ -1211,7 +857,7 @@ static int venus_core_trigger_ssr(struct venus_core
>> *core, u32 trigger_type)
>> if (ret)
>> return ret;
>> - return venus_iface_cmdq_write(hdev, &pkt, false);
>> + return venus_iface_cmdq_write(hdev, &pkt, pkt.hdr.size, false);
>> }
>> static int venus_session_init(struct venus_inst *inst, u32 session_type,
>> @@ -1229,7 +875,7 @@ static int venus_session_init(struct venus_inst
>> *inst, u32 session_type,
>> if (ret)
>> goto err;
>> - ret = venus_iface_cmdq_write(hdev, &pkt, true);
>> + ret = venus_iface_cmdq_write(hdev, &pkt, pkt.shdr.hdr.size, true);
>> if (ret)
>> goto err;
>> @@ -1272,7 +918,7 @@ static int venus_session_flush(struct venus_inst
>> *inst, u32 flush_mode)
>> if (ret)
>> return ret;
>> - return venus_iface_cmdq_write(hdev, &pkt, true);
>> + return venus_iface_cmdq_write(hdev, &pkt, pkt.shdr.hdr.size, true);
>> }
>> static int venus_session_start(struct venus_inst *inst)
>> @@ -1304,7 +950,7 @@ static int venus_session_etb(struct venus_inst *inst,
>> if (ret)
>> return ret;
>> - ret = venus_iface_cmdq_write(hdev, &pkt, false);
>> + ret = venus_iface_cmdq_write(hdev, &pkt, pkt.shdr.hdr.size, false);
>> } else if (session_type == VIDC_SESSION_TYPE_ENC) {
>> struct hfi_session_empty_buffer_uncompressed_plane0_pkt pkt;
>> @@ -1312,7 +958,7 @@ static int venus_session_etb(struct venus_inst *inst,
>> if (ret)
>> return ret;
>> - ret = venus_iface_cmdq_write(hdev, &pkt, false);
>> + ret = venus_iface_cmdq_write(hdev, &pkt, pkt.shdr.hdr.size, false);
>> } else {
>> ret = -EINVAL;
>> }
>> @@ -1331,7 +977,7 @@ static int venus_session_ftb(struct venus_inst *inst,
>> if (ret)
>> return ret;
>> - return venus_iface_cmdq_write(hdev, &pkt, false);
>> + return venus_iface_cmdq_write(hdev, &pkt, pkt.shdr.hdr.size, false);
>> }
>> static int venus_session_set_buffers(struct venus_inst *inst,
>> @@ -1351,7 +997,7 @@ static int venus_session_set_buffers(struct
>> venus_inst *inst,
>> if (ret)
>> return ret;
>> - return venus_iface_cmdq_write(hdev, pkt, false);
>> + return venus_iface_cmdq_write(hdev, pkt, pkt->shdr.hdr.size, false);
>> }
>> static int venus_session_unset_buffers(struct venus_inst *inst,
>> @@ -1371,7 +1017,7 @@ static int venus_session_unset_buffers(struct
>> venus_inst *inst,
>> if (ret)
>> return ret;
>> - return venus_iface_cmdq_write(hdev, pkt, true);
>> + return venus_iface_cmdq_write(hdev, pkt, pkt->shdr.hdr.size, true);
>> }
>> static int venus_session_load_res(struct venus_inst *inst)
>> @@ -1398,7 +1044,7 @@ static int venus_session_parse_seq_hdr(struct
>> venus_inst *inst, u32 seq_hdr,
>> if (ret)
>> return ret;
>> - ret = venus_iface_cmdq_write(hdev, pkt, false);
>> + ret = venus_iface_cmdq_write(hdev, pkt, pkt->shdr.hdr.size, false);
>> if (ret)
>> return ret;
>> @@ -1419,7 +1065,7 @@ static int venus_session_get_seq_hdr(struct
>> venus_inst *inst, u32 seq_hdr,
>> if (ret)
>> return ret;
>> - return venus_iface_cmdq_write(hdev, pkt, false);
>> + return venus_iface_cmdq_write(hdev, pkt, pkt->shdr.hdr.size, false);
>> }
>> static int venus_session_set_property(struct venus_inst *inst, u32
>> ptype,
>> @@ -1438,7 +1084,7 @@ static int venus_session_set_property(struct
>> venus_inst *inst, u32 ptype,
>> if (ret)
>> return ret;
>> - return venus_iface_cmdq_write(hdev, pkt, false);
>> + return venus_iface_cmdq_write(hdev, pkt, pkt->shdr.hdr.size, false);
>> }
>> static int venus_session_get_property(struct venus_inst *inst, u32
>> ptype)
>> @@ -1451,7 +1097,7 @@ static int venus_session_get_property(struct
>> venus_inst *inst, u32 ptype)
>> if (ret)
>> return ret;
>> - return venus_iface_cmdq_write(hdev, &pkt, true);
>> + return venus_iface_cmdq_write(hdev, &pkt, pkt.shdr.hdr.size, true);
>> }
>> static int venus_resume(struct venus_core *core)
>> @@ -1685,7 +1331,11 @@ void venus_hfi_destroy(struct venus_core *core)
>> struct venus_hfi_device *hdev = to_hfi_priv(core);
>> core->priv = NULL;
>> - venus_interface_queues_release(hdev);
>> + mutex_lock(&hdev->lock);
>> + hfi_queue_deinit(hdev->core->dev, &hdev->ifaceq_table, &hdev->sfr,
>> + &hdev->queues[0], &hdev->queues[1],
>> + &hdev->queues[2]);
>> + mutex_unlock(&hdev->lock);
>> mutex_destroy(&hdev->lock);
>> kfree(hdev);
>> core->ops = NULL;
>> @@ -1707,7 +1357,9 @@ int venus_hfi_create(struct venus_core *core)
>> core->priv = hdev;
>> core->ops = &venus_hfi_ops;
>> - ret = venus_interface_queues_init(hdev);
>> + ret = hfi_queue_init(hdev->core->dev, &hdev->ifaceq_table, &hdev->sfr,
>> + &hdev->queues[0], &hdev->queues[1],
>> + &hdev->queues[2], &hdev->core);
>> if (ret)
>> goto err_kfree;
>> @@ -1724,7 +1376,7 @@ void venus_hfi_queues_reinit(struct venus_core
>> *core)
>> {
>> struct venus_hfi_device *hdev = to_hfi_priv(core);
>> struct hfi_queue_table_header *tbl_hdr;
>> - struct iface_queue *queue;
>> + struct iface_q_info *queue;
>> struct hfi_sfr *sfr;
>> unsigned int i;
>> @@ -1733,36 +1385,36 @@ void venus_hfi_queues_reinit(struct venus_core
>> *core)
>> for (i = 0; i < IFACEQ_NUM; i++) {
>> queue = &hdev->queues[i];
>> queue->qhdr =
>> - IFACEQ_GET_QHDR_START_ADDR(hdev->ifaceq_table.kva, i);
>> + IFACEQ_GET_QHDR_START_ADDR(hdev->ifaceq_table.kernel_vaddr, i);
>> venus_set_qhdr_defaults(queue->qhdr);
>> - queue->qhdr->start_addr = queue->qmem.da;
>> + queue->qhdr->qhdr_start_addr = queue->q_array.device_addr;
>> if (i == IFACEQ_CMD_IDX)
>> - queue->qhdr->type |= HFI_HOST_TO_CTRL_CMD_Q;
>> + queue->qhdr->qhdr_type |= HFI_HOST_TO_CTRL_CMD_Q;
>> else if (i == IFACEQ_MSG_IDX)
>> - queue->qhdr->type |= HFI_CTRL_TO_HOST_MSG_Q;
>> + queue->qhdr->qhdr_type |= HFI_CTRL_TO_HOST_MSG_Q;
>> else if (i == IFACEQ_DBG_IDX)
>> - queue->qhdr->type |= HFI_CTRL_TO_HOST_DBG_Q;
>> + queue->qhdr->qhdr_type |= HFI_CTRL_TO_HOST_DBG_Q;
>> }
>> - tbl_hdr = hdev->ifaceq_table.kva;
>> - tbl_hdr->version = 0;
>> - tbl_hdr->size = IFACEQ_TABLE_SIZE;
>> - tbl_hdr->qhdr0_offset = sizeof(struct hfi_queue_table_header);
>> - tbl_hdr->qhdr_size = sizeof(struct hfi_queue_header);
>> - tbl_hdr->num_q = IFACEQ_NUM;
>> - tbl_hdr->num_active_q = IFACEQ_NUM;
>> + tbl_hdr = hdev->ifaceq_table.kernel_vaddr;
>> + tbl_hdr->qtbl_version = 0;
>> + tbl_hdr->qtbl_size = IFACEQ_TABLE_SIZE;
>> + tbl_hdr->qtbl_qhdr0_offset = sizeof(struct hfi_queue_table_header);
>> + tbl_hdr->qtbl_qhdr_size = sizeof(struct hfi_queue_header);
>> + tbl_hdr->qtbl_num_q = IFACEQ_NUM;
>> + tbl_hdr->qtbl_num_active_q = IFACEQ_NUM;
>
> NAK. This is not 'an introduction of helpers'.
Sure, will fix this.
>
>> /*
>> * Set receive request to zero on debug queue as there is no
>> * need of interrupt from video hardware for debug messages
>> */
>> queue = &hdev->queues[IFACEQ_DBG_IDX];
>> - queue->qhdr->rx_req = 0;
>> + queue->qhdr->qhdr_rx_req = 0;
>> - sfr = hdev->sfr.kva;
>> + sfr = hdev->sfr.kernel_vaddr;
>> sfr->buf_size = ALIGNED_SFR_SIZE;
>> /* ensure table and queue header structs are settled in memory */
>
Powered by blists - more mailing lists