diff --git a/drivers/spi/spi-tegra210-quad.c b/drivers/spi/spi-tegra210-quad.c index e2118bd7efc3d3b549117e10e9739e8afcbc43ed..cbd3107dd5708bc952185004889aa47b017df207 100644 --- a/drivers/spi/spi-tegra210-quad.c +++ b/drivers/spi/spi-tegra210-quad.c @@ -22,6 +22,7 @@ #include #include #include +#include #define QSPI_COMMAND1 0x000 #define QSPI_BIT_LENGTH(x) (((x) & 0x1f) << 0) @@ -110,6 +111,9 @@ #define QSPI_DMA_BLK 0x024 #define QSPI_DMA_BLK_SET(x) (((x) & 0xffff) << 0) +#define QSPI_DMA_MEM_ADDRESS 0x028 +#define QSPI_DMA_HI_ADDRESS 0x02c + #define QSPI_TX_FIFO 0x108 #define QSPI_RX_FIFO 0x188 @@ -156,15 +160,19 @@ #define DATA_DIR_RX BIT(1) #define QSPI_DMA_TIMEOUT (msecs_to_jiffies(1000)) -#define DEFAULT_QSPI_DMA_BUF_LEN (64 * 1024) -#define CMD_TRANSFER 0 -#define ADDR_TRANSFER 1 -#define DATA_TRANSFER 2 +#define DEFAULT_QSPI_DMA_BUF_LEN SZ_64K + +enum tegra_qspi_transfer_type { + CMD_TRANSFER = 0, + ADDR_TRANSFER = 1, + DUMMY_TRANSFER = 2, + DATA_TRANSFER = 3 +}; struct tegra_qspi_soc_data { - bool has_dma; bool cmb_xfer_capable; bool supports_tpm; + bool has_ext_dma; unsigned int cs_count; }; @@ -399,9 +407,6 @@ tegra_qspi_read_rx_fifo_to_client_rxbuf(struct tegra_qspi *tqspi, struct spi_tra static void tegra_qspi_copy_client_txbuf_to_qspi_txbuf(struct tegra_qspi *tqspi, struct spi_transfer *t) { - dma_sync_single_for_cpu(tqspi->dev, tqspi->tx_dma_phys, - tqspi->dma_buf_size, DMA_TO_DEVICE); - /* * In packed mode, each word in FIFO may contain multiple packets * based on bits per word. So all bytes in each FIFO word are valid. @@ -434,17 +439,11 @@ tegra_qspi_copy_client_txbuf_to_qspi_txbuf(struct tegra_qspi *tqspi, struct spi_ tqspi->cur_tx_pos += write_bytes; } - - dma_sync_single_for_device(tqspi->dev, tqspi->tx_dma_phys, - tqspi->dma_buf_size, DMA_TO_DEVICE); } static void tegra_qspi_copy_qspi_rxbuf_to_client_rxbuf(struct tegra_qspi *tqspi, struct spi_transfer *t) { - dma_sync_single_for_cpu(tqspi->dev, tqspi->rx_dma_phys, - tqspi->dma_buf_size, DMA_FROM_DEVICE); - if (tqspi->is_packed) { tqspi->cur_rx_pos += tqspi->curr_dma_words * tqspi->bytes_per_word; } else { @@ -470,9 +469,6 @@ tegra_qspi_copy_qspi_rxbuf_to_client_rxbuf(struct tegra_qspi *tqspi, struct spi_ tqspi->cur_rx_pos += read_bytes; } - - dma_sync_single_for_device(tqspi->dev, tqspi->rx_dma_phys, - tqspi->dma_buf_size, DMA_FROM_DEVICE); } static void tegra_qspi_dma_complete(void *args) @@ -600,13 +596,16 @@ static void tegra_qspi_dma_unmap_xfer(struct tegra_qspi *tqspi, struct spi_trans len = DIV_ROUND_UP(tqspi->curr_dma_words * tqspi->bytes_per_word, 4) * 4; - dma_unmap_single(tqspi->dev, t->tx_dma, len, DMA_TO_DEVICE); - dma_unmap_single(tqspi->dev, t->rx_dma, len, DMA_FROM_DEVICE); + if (t->tx_buf) + dma_unmap_single(tqspi->dev, t->tx_dma, len, DMA_TO_DEVICE); + if (t->rx_buf) + dma_unmap_single(tqspi->dev, t->rx_dma, len, DMA_FROM_DEVICE); } static int tegra_qspi_start_dma_based_transfer(struct tegra_qspi *tqspi, struct spi_transfer *t) { struct dma_slave_config dma_sconfig = { 0 }; + dma_addr_t rx_dma_phys, tx_dma_phys; unsigned int len; u8 dma_burst; int ret = 0; @@ -629,60 +628,84 @@ static int tegra_qspi_start_dma_based_transfer(struct tegra_qspi *tqspi, struct len = tqspi->curr_dma_words * 4; /* set attention level based on length of transfer */ - val = 0; - if (len & 0xf) { - val |= QSPI_TX_TRIG_1 | QSPI_RX_TRIG_1; - dma_burst = 1; - } else if (((len) >> 4) & 0x1) { - val |= QSPI_TX_TRIG_4 | QSPI_RX_TRIG_4; - dma_burst = 4; - } else { - val |= QSPI_TX_TRIG_8 | QSPI_RX_TRIG_8; - dma_burst = 8; + if (tqspi->soc_data->has_ext_dma) { + val = 0; + if (len & 0xf) { + val |= QSPI_TX_TRIG_1 | QSPI_RX_TRIG_1; + dma_burst = 1; + } else if (((len) >> 4) & 0x1) { + val |= QSPI_TX_TRIG_4 | QSPI_RX_TRIG_4; + dma_burst = 4; + } else { + val |= QSPI_TX_TRIG_8 | QSPI_RX_TRIG_8; + dma_burst = 8; + } + + tegra_qspi_writel(tqspi, val, QSPI_DMA_CTL); } - tegra_qspi_writel(tqspi, val, QSPI_DMA_CTL); tqspi->dma_control_reg = val; dma_sconfig.device_fc = true; + if (tqspi->cur_direction & DATA_DIR_TX) { - dma_sconfig.dst_addr = tqspi->phys + QSPI_TX_FIFO; - dma_sconfig.dst_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES; - dma_sconfig.dst_maxburst = dma_burst; - ret = dmaengine_slave_config(tqspi->tx_dma_chan, &dma_sconfig); - if (ret < 0) { - dev_err(tqspi->dev, "failed DMA slave config: %d\n", ret); - return ret; - } + if (tqspi->tx_dma_chan) { + dma_sconfig.dst_addr = tqspi->phys + QSPI_TX_FIFO; + dma_sconfig.dst_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES; + dma_sconfig.dst_maxburst = dma_burst; + ret = dmaengine_slave_config(tqspi->tx_dma_chan, &dma_sconfig); + if (ret < 0) { + dev_err(tqspi->dev, "failed DMA slave config: %d\n", ret); + return ret; + } - tegra_qspi_copy_client_txbuf_to_qspi_txbuf(tqspi, t); - ret = tegra_qspi_start_tx_dma(tqspi, t, len); - if (ret < 0) { - dev_err(tqspi->dev, "failed to starting TX DMA: %d\n", ret); - return ret; + tegra_qspi_copy_client_txbuf_to_qspi_txbuf(tqspi, t); + ret = tegra_qspi_start_tx_dma(tqspi, t, len); + if (ret < 0) { + dev_err(tqspi->dev, "failed to starting TX DMA: %d\n", ret); + return ret; + } + } else { + if (tqspi->is_packed) + tx_dma_phys = t->tx_dma; + else + tx_dma_phys = tqspi->tx_dma_phys; + tegra_qspi_copy_client_txbuf_to_qspi_txbuf(tqspi, t); + tegra_qspi_writel(tqspi, lower_32_bits(tx_dma_phys), + QSPI_DMA_MEM_ADDRESS); + tegra_qspi_writel(tqspi, (upper_32_bits(tx_dma_phys) & 0xff), + QSPI_DMA_HI_ADDRESS); } } if (tqspi->cur_direction & DATA_DIR_RX) { - dma_sconfig.src_addr = tqspi->phys + QSPI_RX_FIFO; - dma_sconfig.src_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES; - dma_sconfig.src_maxburst = dma_burst; - ret = dmaengine_slave_config(tqspi->rx_dma_chan, &dma_sconfig); - if (ret < 0) { - dev_err(tqspi->dev, "failed DMA slave config: %d\n", ret); - return ret; - } - - dma_sync_single_for_device(tqspi->dev, tqspi->rx_dma_phys, - tqspi->dma_buf_size, - DMA_FROM_DEVICE); + if (tqspi->rx_dma_chan) { + dma_sconfig.src_addr = tqspi->phys + QSPI_RX_FIFO; + dma_sconfig.src_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES; + dma_sconfig.src_maxburst = dma_burst; + ret = dmaengine_slave_config(tqspi->rx_dma_chan, &dma_sconfig); + if (ret < 0) { + dev_err(tqspi->dev, "failed DMA slave config: %d\n", ret); + return ret; + } - ret = tegra_qspi_start_rx_dma(tqspi, t, len); - if (ret < 0) { - dev_err(tqspi->dev, "failed to start RX DMA: %d\n", ret); - if (tqspi->cur_direction & DATA_DIR_TX) - dmaengine_terminate_all(tqspi->tx_dma_chan); - return ret; + ret = tegra_qspi_start_rx_dma(tqspi, t, len); + if (ret < 0) { + dev_err(tqspi->dev, "failed to start RX DMA: %d\n", ret); + if (tqspi->cur_direction & DATA_DIR_TX) + dmaengine_terminate_all(tqspi->tx_dma_chan); + return ret; + } + } else { + if (tqspi->is_packed) + rx_dma_phys = t->rx_dma; + else + rx_dma_phys = tqspi->rx_dma_phys; + + tegra_qspi_writel(tqspi, lower_32_bits(rx_dma_phys), + QSPI_DMA_MEM_ADDRESS); + tegra_qspi_writel(tqspi, (upper_32_bits(rx_dma_phys) & 0xff), + QSPI_DMA_HI_ADDRESS); } } @@ -721,9 +744,6 @@ static int tegra_qspi_start_cpu_based_transfer(struct tegra_qspi *qspi, struct s static void tegra_qspi_deinit_dma(struct tegra_qspi *tqspi) { - if (!tqspi->soc_data->has_dma) - return; - if (tqspi->tx_dma_buf) { dma_free_coherent(tqspi->dev, tqspi->dma_buf_size, tqspi->tx_dma_buf, tqspi->tx_dma_phys); @@ -754,16 +774,29 @@ static int tegra_qspi_init_dma(struct tegra_qspi *tqspi) u32 *dma_buf; int err; - if (!tqspi->soc_data->has_dma) - return 0; + if (tqspi->soc_data->has_ext_dma) { + dma_chan = dma_request_chan(tqspi->dev, "rx"); + if (IS_ERR(dma_chan)) { + err = PTR_ERR(dma_chan); + goto err_out; + } - dma_chan = dma_request_chan(tqspi->dev, "rx"); - if (IS_ERR(dma_chan)) { - err = PTR_ERR(dma_chan); - goto err_out; - } + tqspi->rx_dma_chan = dma_chan; + + dma_chan = dma_request_chan(tqspi->dev, "tx"); + if (IS_ERR(dma_chan)) { + err = PTR_ERR(dma_chan); + goto err_out; + } - tqspi->rx_dma_chan = dma_chan; + tqspi->tx_dma_chan = dma_chan; + } else { + if (!device_iommu_mapped(tqspi->dev)) { + dev_warn(tqspi->dev, + "IOMMU not enabled in device-tree, falling back to PIO mode\n"); + return 0; + } + } dma_buf = dma_alloc_coherent(tqspi->dev, tqspi->dma_buf_size, &dma_phys, GFP_KERNEL); if (!dma_buf) { @@ -774,14 +807,6 @@ static int tegra_qspi_init_dma(struct tegra_qspi *tqspi) tqspi->rx_dma_buf = dma_buf; tqspi->rx_dma_phys = dma_phys; - dma_chan = dma_request_chan(tqspi->dev, "tx"); - if (IS_ERR(dma_chan)) { - err = PTR_ERR(dma_chan); - goto err_out; - } - - tqspi->tx_dma_chan = dma_chan; - dma_buf = dma_alloc_coherent(tqspi->dev, tqspi->dma_buf_size, &dma_phys, GFP_KERNEL); if (!dma_buf) { err = -ENOMEM; @@ -814,6 +839,7 @@ static u32 tegra_qspi_setup_transfer_one(struct spi_device *spi, struct spi_tran u32 command1, command2, speed = t->speed_hz; u8 bits_per_word = t->bits_per_word; u32 tx_tap = 0, rx_tap = 0; + unsigned long flags; int req_mode; if (!has_acpi_companion(tqspi->dev) && speed != tqspi->cur_speed) { @@ -821,10 +847,12 @@ static u32 tegra_qspi_setup_transfer_one(struct spi_device *spi, struct spi_tran tqspi->cur_speed = speed; } + spin_lock_irqsave(&tqspi->lock, flags); tqspi->cur_pos = 0; tqspi->cur_rx_pos = 0; tqspi->cur_tx_pos = 0; tqspi->curr_xfer = t; + spin_unlock_irqrestore(&tqspi->lock, flags); if (is_first_of_msg) { tegra_qspi_mask_clear_irq(tqspi); @@ -1133,6 +1161,7 @@ static int tegra_qspi_combined_seq_xfer(struct tegra_qspi *tqspi, u32 address_value = 0; u32 cmd_config = 0, addr_config = 0; u8 cmd_value = 0, val = 0; + unsigned long flags; /* Enable Combined sequence mode */ val = tegra_qspi_readl(tqspi, QSPI_GLOBAL_CONFIG); @@ -1159,6 +1188,13 @@ static int tegra_qspi_combined_seq_xfer(struct tegra_qspi *tqspi, xfer->len); address_value = *((const u32 *)(xfer->tx_buf)); break; + case DUMMY_TRANSFER: + if (xfer->dummy_data) { + tqspi->dummy_cycles = xfer->len * 8 / xfer->tx_nbits; + break; + } + transfer_phase++; + fallthrough; case DATA_TRANSFER: /* Program Command, Address value in register */ tegra_qspi_writel(tqspi, cmd_value, QSPI_CMB_SEQ_CMD); @@ -1225,17 +1261,21 @@ static int tegra_qspi_combined_seq_xfer(struct tegra_qspi *tqspi, goto exit; } msg->actual_length += xfer->len; - tqspi->curr_xfer = NULL; if (!xfer->cs_change && transfer_phase == DATA_TRANSFER) { tegra_qspi_transfer_end(spi); spi_transfer_delay_exec(xfer); } + spin_lock_irqsave(&tqspi->lock, flags); + tqspi->curr_xfer = NULL; + spin_unlock_irqrestore(&tqspi->lock, flags); transfer_phase++; } ret = 0; exit: + spin_lock_irqsave(&tqspi->lock, flags); tqspi->curr_xfer = NULL; + spin_unlock_irqrestore(&tqspi->lock, flags); msg->status = ret; return ret; @@ -1248,6 +1288,7 @@ static int tegra_qspi_non_combined_seq_xfer(struct tegra_qspi *tqspi, struct spi_transfer *transfer; bool is_first_msg = true; int ret = 0, val = 0; + unsigned long flags; msg->status = 0; msg->actual_length = 0; @@ -1328,7 +1369,9 @@ static int tegra_qspi_non_combined_seq_xfer(struct tegra_qspi *tqspi, msg->actual_length += xfer->len + dummy_bytes; complete_xfer: + spin_lock_irqsave(&tqspi->lock, flags); tqspi->curr_xfer = NULL; + spin_unlock_irqrestore(&tqspi->lock, flags); if (ret < 0) { tegra_qspi_transfer_end(spi); @@ -1365,7 +1408,9 @@ static bool tegra_qspi_validate_cmb_seq(struct tegra_qspi *tqspi, list_for_each_entry(xfer, &msg->transfers, transfer_list) { transfer_count++; } - if (!tqspi->soc_data->cmb_xfer_capable || transfer_count != 3) + if (!tqspi->soc_data->cmb_xfer_capable) + return false; + if (transfer_count > 4 || transfer_count < 3) return false; xfer = list_first_entry(&msg->transfers, typeof(*xfer), transfer_list); @@ -1375,7 +1420,14 @@ static bool tegra_qspi_validate_cmb_seq(struct tegra_qspi *tqspi, if (xfer->len > 4 || xfer->len < 3) return false; xfer = list_next_entry(xfer, transfer_list); - if (!tqspi->soc_data->has_dma && xfer->len > (QSPI_FIFO_DEPTH << 2)) + if (transfer_count == 4) { + if (xfer->dummy_data != 1) + return false; + if ((xfer->len * 8 / xfer->tx_nbits) > QSPI_DUMMY_CYCLES_MAX) + return false; + xfer = list_next_entry(xfer, transfer_list); + } + if (!tqspi->soc_data->has_ext_dma && xfer->len > (QSPI_FIFO_DEPTH << 2)) return false; return true; @@ -1399,10 +1451,16 @@ static int tegra_qspi_transfer_one_message(struct spi_master *master, static irqreturn_t handle_cpu_based_xfer(struct tegra_qspi *tqspi) { - struct spi_transfer *t = tqspi->curr_xfer; + struct spi_transfer *t; unsigned long flags; spin_lock_irqsave(&tqspi->lock, flags); + t = tqspi->curr_xfer; + + if (!t) { + spin_unlock_irqrestore(&tqspi->lock, flags); + return IRQ_HANDLED; + } if (tqspi->tx_status || tqspi->rx_status) { tegra_qspi_handle_error(tqspi); @@ -1426,51 +1484,60 @@ static irqreturn_t handle_cpu_based_xfer(struct tegra_qspi *tqspi) tegra_qspi_calculate_curr_xfer_param(tqspi, t); tegra_qspi_start_cpu_based_transfer(tqspi, t); exit: + tqspi->curr_xfer = NULL; spin_unlock_irqrestore(&tqspi->lock, flags); return IRQ_HANDLED; } static irqreturn_t handle_dma_based_xfer(struct tegra_qspi *tqspi) { - struct spi_transfer *t = tqspi->curr_xfer; + struct spi_transfer *t; unsigned int total_fifo_words; unsigned long flags; long wait_status; - int err = 0; + int num_errors = 0; if (tqspi->cur_direction & DATA_DIR_TX) { if (tqspi->tx_status) { - dmaengine_terminate_all(tqspi->tx_dma_chan); - err += 1; - } else { + if (tqspi->tx_dma_chan) + dmaengine_terminate_all(tqspi->tx_dma_chan); + num_errors++; + } else if (tqspi->tx_dma_chan) { wait_status = wait_for_completion_interruptible_timeout( &tqspi->tx_dma_complete, QSPI_DMA_TIMEOUT); if (wait_status <= 0) { dmaengine_terminate_all(tqspi->tx_dma_chan); dev_err(tqspi->dev, "failed TX DMA transfer\n"); - err += 1; + num_errors++; } } } if (tqspi->cur_direction & DATA_DIR_RX) { if (tqspi->rx_status) { - dmaengine_terminate_all(tqspi->rx_dma_chan); - err += 2; - } else { + if (tqspi->rx_dma_chan) + dmaengine_terminate_all(tqspi->rx_dma_chan); + num_errors++; + } else if (tqspi->rx_dma_chan) { wait_status = wait_for_completion_interruptible_timeout( &tqspi->rx_dma_complete, QSPI_DMA_TIMEOUT); if (wait_status <= 0) { dmaengine_terminate_all(tqspi->rx_dma_chan); dev_err(tqspi->dev, "failed RX DMA transfer\n"); - err += 2; + num_errors++; } } } spin_lock_irqsave(&tqspi->lock, flags); + t = tqspi->curr_xfer; + + if (!t) { + spin_unlock_irqrestore(&tqspi->lock, flags); + return IRQ_HANDLED; + } - if (err) { + if (num_errors) { tegra_qspi_dma_unmap_xfer(tqspi, t); tegra_qspi_handle_error(tqspi); complete(&tqspi->xfer_completion); @@ -1496,9 +1563,9 @@ static irqreturn_t handle_dma_based_xfer(struct tegra_qspi *tqspi) /* continue transfer in current message */ total_fifo_words = tegra_qspi_calculate_curr_xfer_param(tqspi, t); if (total_fifo_words > QSPI_FIFO_DEPTH) - err = tegra_qspi_start_dma_based_transfer(tqspi, t); + num_errors = tegra_qspi_start_dma_based_transfer(tqspi, t); else - err = tegra_qspi_start_cpu_based_transfer(tqspi, t); + num_errors = tegra_qspi_start_cpu_based_transfer(tqspi, t); exit: spin_unlock_irqrestore(&tqspi->lock, flags); @@ -1508,15 +1575,33 @@ static irqreturn_t handle_dma_based_xfer(struct tegra_qspi *tqspi) static irqreturn_t tegra_qspi_isr_thread(int irq, void *context_data) { struct tegra_qspi *tqspi = context_data; + unsigned long flags; + u32 status; + + /* + * Read transfer status to check if interrupt was triggered by transfer + * completion + */ + status = tegra_qspi_readl(tqspi, QSPI_TRANS_STATUS); /* * Occasionally the IRQ thread takes a long time to wake up (usually * when the CPU that it's running on is excessively busy) and we have * already reached the timeout before and cleaned up the timed out * transfer. Avoid any processing in that case and bail out early. + * + * If no transfer is in progress, check if this was a real interrupt + * that the timeout handler already processed, or a spurious one. */ - if (tqspi->curr_xfer == NULL) - return IRQ_NONE; + spin_lock_irqsave(&tqspi->lock, flags); + if (!tqspi->curr_xfer) { + spin_unlock_irqrestore(&tqspi->lock, flags); + /* Spurious interrupt - transfer not ready */ + if (!(status & QSPI_RDY)) + return IRQ_NONE; + /* Real interrupt, already handled by timeout path */ + return IRQ_HANDLED; + } tqspi->status_reg = tegra_qspi_readl(tqspi, QSPI_FIFO_STATUS); @@ -1527,7 +1612,14 @@ static irqreturn_t tegra_qspi_isr_thread(int irq, void *context_data) tqspi->rx_status = tqspi->status_reg & (QSPI_RX_FIFO_OVF | QSPI_RX_FIFO_UNF); tegra_qspi_mask_clear_irq(tqspi); + spin_unlock_irqrestore(&tqspi->lock, flags); + /* + * Lock is released here but handlers safely re-check curr_xfer under + * lock before dereferencing. + * DMA handler also needs to sleep in wait_for_completion_*(), which + * cannot be done while holding spinlock. + */ if (!tqspi->is_curr_dma_xfer) return handle_cpu_based_xfer(tqspi); @@ -1535,28 +1627,28 @@ static irqreturn_t tegra_qspi_isr_thread(int irq, void *context_data) } static struct tegra_qspi_soc_data tegra210_qspi_soc_data = { - .has_dma = true, + .has_ext_dma = true, .cmb_xfer_capable = false, .supports_tpm = false, .cs_count = 1, }; static struct tegra_qspi_soc_data tegra186_qspi_soc_data = { - .has_dma = true, + .has_ext_dma = true, .cmb_xfer_capable = true, .supports_tpm = false, .cs_count = 1, }; static struct tegra_qspi_soc_data tegra234_qspi_soc_data = { - .has_dma = false, + .has_ext_dma = false, .cmb_xfer_capable = true, .supports_tpm = true, .cs_count = 1, }; static struct tegra_qspi_soc_data tegra241_qspi_soc_data = { - .has_dma = false, + .has_ext_dma = true, .cmb_xfer_capable = true, .supports_tpm = true, .cs_count = 4,