net: ipa: rename a few endpoint config data types
authorAlex Elder <elder@linaro.org>
Thu, 19 May 2022 15:12:16 +0000 (10:12 -0500)
committerDavid S. Miller <davem@davemloft.net>
Fri, 20 May 2022 10:12:24 +0000 (11:12 +0100)
Rename the just-moved data structure types to drop the "_data"
suffix, to make it more obvious they are no longer meant to be used
just as read-only initialization data.  Rename the fields and
variables of these types to use "config" instead of "data" in the
name.  This is another small step meant to facilitate review.

Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
drivers/net/ipa/ipa_data.h
drivers/net/ipa/ipa_endpoint.c
drivers/net/ipa/ipa_endpoint.h
drivers/net/ipa/ipa_modem.c

index d611b5e96497c387ee7b78e6a1cbd4d9c635f294..e15eb3cd3e333bc925f3c3e8bd91d63151f85577 100644 (file)
@@ -98,7 +98,7 @@ struct gsi_channel_data {
 /**
  * struct ipa_endpoint_data - IPA endpoint configuration data
  * @filter_support:    whether endpoint supports filtering
- * @config:            hardware configuration (see above)
+ * @config:            hardware configuration
  *
  * Not all endpoints support the IPA filtering capability.  A filter table
  * defines the filters to apply for those endpoints that support it.  The
@@ -106,12 +106,12 @@ struct gsi_channel_data {
  * for non-AP endpoints.  For this reason we define *all* endpoints used
  * in the system, and indicate whether they support filtering.
  *
- * The remaining endpoint configuration data applies only to AP endpoints.
+ * The remaining endpoint configuration data specifies default hardware
+ * configuration values that apply only to AP endpoints.
  */
 struct ipa_endpoint_data {
        bool filter_support;
-       /* Everything else is specified only for AP endpoints */
-       struct ipa_endpoint_config_data config;
+       struct ipa_endpoint_config config;
 };
 
 /**
index 55322800ba58e22953280110b72804a303a8d92e..361f9e17e8697c48bcc60c6a5c290daa4a1f8114 100644 (file)
@@ -333,7 +333,7 @@ static void ipa_endpoint_suspend_aggr(struct ipa_endpoint *endpoint)
 {
        struct ipa *ipa = endpoint->ipa;
 
-       if (!endpoint->data->aggregation)
+       if (!endpoint->config->aggregation)
                return;
 
        /* Nothing to do if the endpoint doesn't have aggregation open */
@@ -453,7 +453,7 @@ static void ipa_endpoint_init_cfg(struct ipa_endpoint *endpoint)
        u32 val = 0;
 
        /* FRAG_OFFLOAD_EN is 0 */
-       if (endpoint->data->checksum) {
+       if (endpoint->config->checksum) {
                enum ipa_version version = endpoint->ipa->version;
 
                if (endpoint->toward_ipa) {
@@ -502,7 +502,7 @@ ipa_qmap_header_size(enum ipa_version version, struct ipa_endpoint *endpoint)
        u32 header_size = sizeof(struct rmnet_map_header);
 
        /* Without checksum offload, we just have the MAP header */
-       if (!endpoint->data->checksum)
+       if (!endpoint->config->checksum)
                return header_size;
 
        if (version < IPA_VERSION_4_5) {
@@ -544,7 +544,7 @@ static void ipa_endpoint_init_hdr(struct ipa_endpoint *endpoint)
        struct ipa *ipa = endpoint->ipa;
        u32 val = 0;
 
-       if (endpoint->data->qmap) {
+       if (endpoint->config->qmap) {
                enum ipa_version version = ipa->version;
                size_t header_size;
 
@@ -583,11 +583,11 @@ static void ipa_endpoint_init_hdr(struct ipa_endpoint *endpoint)
 static void ipa_endpoint_init_hdr_ext(struct ipa_endpoint *endpoint)
 {
        u32 offset = IPA_REG_ENDP_INIT_HDR_EXT_N_OFFSET(endpoint->endpoint_id);
-       u32 pad_align = endpoint->data->rx.pad_align;
+       u32 pad_align = endpoint->config->rx.pad_align;
        struct ipa *ipa = endpoint->ipa;
        u32 val = 0;
 
-       if (endpoint->data->qmap) {
+       if (endpoint->config->qmap) {
                /* We have a header, so we must specify its endianness */
                val |= HDR_ENDIANNESS_FMASK;    /* big endian */
 
@@ -615,7 +615,7 @@ static void ipa_endpoint_init_hdr_ext(struct ipa_endpoint *endpoint)
         */
        if (ipa->version >= IPA_VERSION_4_5) {
                /* HDR_TOTAL_LEN_OR_PAD_OFFSET is 0, so MSB is 0 */
-               if (endpoint->data->qmap && !endpoint->toward_ipa) {
+               if (endpoint->config->qmap && !endpoint->toward_ipa) {
                        u32 offset;
 
                        offset = offsetof(struct rmnet_map_header, pkt_len);
@@ -640,7 +640,7 @@ static void ipa_endpoint_init_hdr_metadata_mask(struct ipa_endpoint *endpoint)
        offset = IPA_REG_ENDP_INIT_HDR_METADATA_MASK_N_OFFSET(endpoint_id);
 
        /* Note that HDR_ENDIANNESS indicates big endian header fields */
-       if (endpoint->data->qmap)
+       if (endpoint->config->qmap)
                val = (__force u32)cpu_to_be32(IPA_ENDPOINT_QMAP_METADATA_MASK);
 
        iowrite32(val, endpoint->ipa->reg_virt + offset);
@@ -654,8 +654,8 @@ static void ipa_endpoint_init_mode(struct ipa_endpoint *endpoint)
        if (!endpoint->toward_ipa)
                return;         /* Register not valid for RX endpoints */
 
-       if (endpoint->data->dma_mode) {
-               enum ipa_endpoint_name name = endpoint->data->dma_endpoint;
+       if (endpoint->config->dma_mode) {
+               enum ipa_endpoint_name name = endpoint->config->dma_endpoint;
                u32 dma_endpoint_id;
 
                dma_endpoint_id = endpoint->ipa->name_map[name]->endpoint_id;
@@ -741,18 +741,18 @@ static void ipa_endpoint_init_aggr(struct ipa_endpoint *endpoint)
        enum ipa_version version = endpoint->ipa->version;
        u32 val = 0;
 
-       if (endpoint->data->aggregation) {
+       if (endpoint->config->aggregation) {
                if (!endpoint->toward_ipa) {
-                       const struct ipa_endpoint_rx_data *rx_data;
+                       const struct ipa_endpoint_rx *rx_config;
                        u32 buffer_size;
                        bool close_eof;
                        u32 limit;
 
-                       rx_data = &endpoint->data->rx;
+                       rx_config = &endpoint->config->rx;
                        val |= u32_encode_bits(IPA_ENABLE_AGGR, AGGR_EN_FMASK);
                        val |= u32_encode_bits(IPA_GENERIC, AGGR_TYPE_FMASK);
 
-                       buffer_size = rx_data->buffer_size;
+                       buffer_size = rx_config->buffer_size;
                        limit = ipa_aggr_size_kb(buffer_size - NET_SKB_PAD);
                        val |= aggr_byte_limit_encoded(version, limit);
 
@@ -761,7 +761,7 @@ static void ipa_endpoint_init_aggr(struct ipa_endpoint *endpoint)
 
                        /* AGGR_PKT_LIMIT is 0 (unlimited) */
 
-                       close_eof = rx_data->aggr_close_eof;
+                       close_eof = rx_config->aggr_close_eof;
                        val |= aggr_sw_eof_active_encoded(version, close_eof);
                } else {
                        val |= u32_encode_bits(IPA_ENABLE_DEAGGR,
@@ -947,7 +947,7 @@ static void ipa_endpoint_init_rsrc_grp(struct ipa_endpoint *endpoint)
        struct ipa *ipa = endpoint->ipa;
        u32 val;
 
-       val = rsrc_grp_encoded(ipa->version, endpoint->data->resource_group);
+       val = rsrc_grp_encoded(ipa->version, endpoint->config->resource_group);
        iowrite32(val, ipa->reg_virt + offset);
 }
 
@@ -960,10 +960,10 @@ static void ipa_endpoint_init_seq(struct ipa_endpoint *endpoint)
                return;         /* Register not valid for RX endpoints */
 
        /* Low-order byte configures primary packet processing */
-       val |= u32_encode_bits(endpoint->data->tx.seq_type, SEQ_TYPE_FMASK);
+       val |= u32_encode_bits(endpoint->config->tx.seq_type, SEQ_TYPE_FMASK);
 
        /* Second byte configures replicated packet processing */
-       val |= u32_encode_bits(endpoint->data->tx.seq_rep_type,
+       val |= u32_encode_bits(endpoint->config->tx.seq_rep_type,
                               SEQ_REP_TYPE_FMASK);
 
        iowrite32(val, endpoint->ipa->reg_virt + offset);
@@ -1021,13 +1021,13 @@ static void ipa_endpoint_status(struct ipa_endpoint *endpoint)
 
        offset = IPA_REG_ENDP_STATUS_N_OFFSET(endpoint_id);
 
-       if (endpoint->data->status_enable) {
+       if (endpoint->config->status_enable) {
                val |= STATUS_EN_FMASK;
                if (endpoint->toward_ipa) {
                        enum ipa_endpoint_name name;
                        u32 status_endpoint_id;
 
-                       name = endpoint->data->tx.status_endpoint;
+                       name = endpoint->config->tx.status_endpoint;
                        status_endpoint_id = ipa->name_map[name]->endpoint_id;
 
                        val |= u32_encode_bits(status_endpoint_id,
@@ -1051,7 +1051,7 @@ static int ipa_endpoint_replenish_one(struct ipa_endpoint *endpoint,
        u32 len;
        int ret;
 
-       buffer_size = endpoint->data->rx.buffer_size;
+       buffer_size = endpoint->config->rx.buffer_size;
        page = dev_alloc_pages(get_order(buffer_size));
        if (!page)
                return -ENOMEM;
@@ -1168,7 +1168,7 @@ static void ipa_endpoint_skb_copy(struct ipa_endpoint *endpoint,
 static bool ipa_endpoint_skb_build(struct ipa_endpoint *endpoint,
                                   struct page *page, u32 len)
 {
-       u32 buffer_size = endpoint->data->rx.buffer_size;
+       u32 buffer_size = endpoint->config->rx.buffer_size;
        struct sk_buff *skb;
 
        /* Nothing to do if there's no netdev */
@@ -1275,7 +1275,7 @@ static bool ipa_endpoint_status_drop(struct ipa_endpoint *endpoint,
 static void ipa_endpoint_status_parse(struct ipa_endpoint *endpoint,
                                      struct page *page, u32 total_len)
 {
-       u32 buffer_size = endpoint->data->rx.buffer_size;
+       u32 buffer_size = endpoint->config->rx.buffer_size;
        void *data = page_address(page) + NET_SKB_PAD;
        u32 unused = buffer_size - total_len;
        u32 resid = total_len;
@@ -1305,10 +1305,10 @@ static void ipa_endpoint_status_parse(struct ipa_endpoint *endpoint,
                 * And if checksum offload is enabled a trailer containing
                 * computed checksum information will be appended.
                 */
-               align = endpoint->data->rx.pad_align ? : 1;
+               align = endpoint->config->rx.pad_align ? : 1;
                len = le16_to_cpu(status->pkt_len);
                len = sizeof(*status) + ALIGN(len, align);
-               if (endpoint->data->checksum)
+               if (endpoint->config->checksum)
                        len += sizeof(struct rmnet_map_dl_csum_trailer);
 
                if (!ipa_endpoint_status_drop(endpoint, status)) {
@@ -1352,7 +1352,7 @@ static void ipa_endpoint_rx_complete(struct ipa_endpoint *endpoint,
 
        /* Parse or build a socket buffer using the actual received length */
        page = trans->data;
-       if (endpoint->data->status_enable)
+       if (endpoint->config->status_enable)
                ipa_endpoint_status_parse(endpoint, page, trans->len);
        else if (ipa_endpoint_skb_build(endpoint, page, trans->len))
                trans->data = NULL;     /* Pages have been consumed */
@@ -1386,7 +1386,7 @@ void ipa_endpoint_trans_release(struct ipa_endpoint *endpoint,
                struct page *page = trans->data;
 
                if (page) {
-                       u32 buffer_size = endpoint->data->rx.buffer_size;
+                       u32 buffer_size = endpoint->config->rx.buffer_size;
 
                        __free_pages(page, get_order(buffer_size));
                }
@@ -1520,7 +1520,7 @@ static void ipa_endpoint_reset(struct ipa_endpoint *endpoint)
         * All other cases just need to reset the underlying GSI channel.
         */
        special = ipa->version < IPA_VERSION_4_0 && !endpoint->toward_ipa &&
-                       endpoint->data->aggregation;
+                       endpoint->config->aggregation;
        if (special && ipa_endpoint_aggr_active(endpoint))
                ret = ipa_endpoint_reset_rx_aggr(endpoint);
        else
@@ -1835,7 +1835,7 @@ static void ipa_endpoint_init_one(struct ipa *ipa, enum ipa_endpoint_name name,
        endpoint->channel_id = data->channel_id;
        endpoint->endpoint_id = data->endpoint_id;
        endpoint->toward_ipa = data->toward_ipa;
-       endpoint->data = &data->endpoint.config;
+       endpoint->config = &data->endpoint.config;
 
        ipa->initialized |= BIT(endpoint->endpoint_id);
 }
index 85fe15b5d983e9bad1c721da454da98807525c05..e8d1300a600220268ed42acea762fda2b596dc2b 100644 (file)
@@ -41,7 +41,7 @@ enum ipa_endpoint_name {
 #define IPA_ENDPOINT_MAX               32      /* Max supported by driver */
 
 /**
- * struct ipa_endpoint_tx_data - configuration data for TX endpoints
+ * struct ipa_endpoint_tx - Endpoint configuration for TX endpoints
  * @seq_type:          primary packet processing sequencer type
  * @seq_rep_type:      sequencer type for replication processing
  * @status_endpoint:   endpoint to which status elements are sent
@@ -49,17 +49,17 @@ enum ipa_endpoint_name {
  * The @status_endpoint is only valid if the endpoint's @status_enable
  * flag is set.
  */
-struct ipa_endpoint_tx_data {
+struct ipa_endpoint_tx {
        enum ipa_seq_type seq_type;
        enum ipa_seq_rep_type seq_rep_type;
        enum ipa_endpoint_name status_endpoint;
 };
 
 /**
- * struct ipa_endpoint_rx_data - configuration data for RX endpoints
- * @buffer_size: requested receive buffer size (bytes)
- * @pad_align: power-of-2 boundary to which packet payload is aligned
- * @aggr_close_eof: whether aggregation closes on end-of-frame
+ * struct ipa_endpoint_rx - Endpoint configuration for RX endpoints
+ * @buffer_size:       requested receive buffer size (bytes)
+ * @pad_align:         power-of-2 boundary to which packet payload is aligned
+ * @aggr_close_eof:    whether aggregation closes on end-of-frame
  *
  * With each packet it transfers, the IPA hardware can perform certain
  * transformations of its packet data.  One of these is adding pad bytes
@@ -70,14 +70,14 @@ struct ipa_endpoint_tx_data {
  * certain criteria are met.  One of those criteria is the sender indicating
  * a "frame" consisting of several transfers has ended.
  */
-struct ipa_endpoint_rx_data {
+struct ipa_endpoint_rx {
        u32 buffer_size;
        u32 pad_align;
        bool aggr_close_eof;
 };
 
 /**
- * struct ipa_endpoint_config_data - IPA endpoint hardware configuration
+ * struct ipa_endpoint_config - IPA endpoint hardware configuration
  * @resource_group:    resource group to assign endpoint to
  * @checksum:          whether checksum offload is enabled
  * @qmap:              whether endpoint uses QMAP protocol
@@ -88,7 +88,7 @@ struct ipa_endpoint_rx_data {
  * @tx:                        TX-specific endpoint information (see above)
  * @rx:                        RX-specific endpoint information (see above)
  */
-struct ipa_endpoint_config_data {
+struct ipa_endpoint_config {
        u32 resource_group;
        bool checksum;
        bool qmap;
@@ -97,8 +97,8 @@ struct ipa_endpoint_config_data {
        bool dma_mode;
        enum ipa_endpoint_name dma_endpoint;
        union {
-               struct ipa_endpoint_tx_data tx;
-               struct ipa_endpoint_rx_data rx;
+               struct ipa_endpoint_tx tx;
+               struct ipa_endpoint_rx rx;
        };
 };
 
@@ -122,7 +122,7 @@ enum ipa_replenish_flag {
  * @channel_id:                GSI channel used by the endpoint
  * @endpoint_id:       IPA endpoint number
  * @toward_ipa:                Endpoint direction (true = TX, false = RX)
- * @data:              Endpoint configuration data
+ * @config:            Default endpoint configuration
  * @trans_tre_max:     Maximum number of TRE descriptors per transaction
  * @evt_ring_id:       GSI event ring used by the endpoint
  * @netdev:            Network device pointer, if endpoint uses one
@@ -136,7 +136,7 @@ struct ipa_endpoint {
        u32 channel_id;
        u32 endpoint_id;
        bool toward_ipa;
-       const struct ipa_endpoint_config_data *data;
+       const struct ipa_endpoint_config *config;
 
        u32 trans_tre_max;
        u32 evt_ring_id;
index dd6464ced2546d619f2807c92ec065702bb78755..7975e324690bbf56af7e52d741dbd26f2d3a1ef9 100644 (file)
@@ -129,7 +129,7 @@ ipa_start_xmit(struct sk_buff *skb, struct net_device *netdev)
                goto err_drop_skb;
 
        endpoint = ipa->name_map[IPA_ENDPOINT_AP_MODEM_TX];
-       if (endpoint->data->qmap && skb->protocol != htons(ETH_P_MAP))
+       if (endpoint->config->qmap && skb->protocol != htons(ETH_P_MAP))
                goto err_drop_skb;
 
        /* The hardware must be powered for us to transmit */